杏吧

In higher eukaryotes, genes are expressed dynamically in complex spatial and temporal patterns, which are progressively refined to set up body plans and define specific cell-types. The information about when and where each gene is to be expressed is encoded in the sequences of promoter-, enhancer- and silencer regions and realized by transcription factor and cofactor proteins. I am presenting our work towards understanding the how this regulatory information is sequence-encoded and how cells utilize this information with transcription factor and cofactor proteins. We characterize regulatory sequences by functional screens in cell lines and by a genome-scale candidate testing approach in developing Drosophila embryos. We then use deep-learning approaches to model the sequence-to-function relationship for enhancers and build synthetic enhancers de novo. We also employ functional screens, mutagenesis, and biochemistry to dissect functions and mechanisms of transcription regulation in flies and mammals.

How come drones are still mainly human controlled and have such limited autonomy? First, drones operate under significant constraints, including limited computational power, energy capacity, and communication bandwidth. Reinforcement Learning fail to maintain optimal performance under such constraints. We propose sequence AI algorithms that significantly improving compute and energy efficiency. Among the key features are rapid onboard responses and adaptability in dynamic environmental changes, robustness to missing inputs, minimization of sensor usage and the ability to use cheaper sensors to greater effect, as well as making possible the use of cheaper hardware while maintaining peak effectiveness. Second issue is the need of communication and cooperation among drones. Distributed AI is known to suffer explosion of communication needs, and this is not available in realistic swarms of drones. We propose a cooperative AI where the agents are lifelong learners. On the go, they are able to update, learn from failures, and become more expert with more experience. This paradigm enables both collaborative AI without explosive communication as well as a great reduction in the required labeled data (teacher), since the agents peer-teach each other. We suggest that these two directions of research will advance us towards true safe autonomy.

Bacteria regulate homeostatic growth by adjusting proteome composition. In Escherichia coli, this coordination is mediated by (p)ppGpp, which couples amino acid supply with ribosome production. We identify a distinct architecture in Bacillus subtilis, where GTP 鈥� not (p)ppGpp 鈥� controls proteome allocation. Translational inhibition results in GTP depletion and suppresses amino acid biosynthesis via feedback inhibition without altering ribosome abundance, decoupling global amino-acid flux from proteome composition. By artificially tuning GTP, we recoupled flux and proteome, restoring growth to maximal levels. The regulated sub-optimality enables a tradeoff to balance growth and stress resilience. Similar GTP-based strategies suggest evolutionary conservation in Firmicutes. These findings call for a revised view of bacterial physiology, where proteome composition and metabolic flux represent distinct regulatory layers.

Signatures of neural computation are thought to be reflected in the coordinated activity of large neural populations. Neuroscience is now flush with measurements of these activity patterns in humans, animal subjects, and large-scale artificial network models. In this talk, I will address an extensively studied, yet unresolved, question: How should we quantify the extent to which two or more neural circuits have 鈥渟imilar鈥� activation patterns? Without an answer to this question, the field has struggled to investigate basic questions about biological variability and individuality, such as: How do neural representations vary across a healthy population? How do differences in neural population activity correlate with behavioral idiosyncrasies and disorders? How similar are computational mechanisms in biological brains and artificial neural networks? In this talk, I will summarize several mathematical methods that quantify similarity in neural representations and demonstrate how they provide early insights into these questions when applied to biological data and artificial networks.

Advances in generative AI are transforming how we discover and engineer therapeutic proteins. Rather than relying on random screening or animal immunisation, these models can now learn the fundamental principles of molecular recognition and design proteins capable of binding virtually any target. This talk will illustrate how such methods are being applied to address two very different biomedical challenges. The first involves designing synthetic proteins that neutralise deadly snake venoms by targeting key neurotoxins and cytotoxins. These de novo binders were created using AI-driven structural generation and achieved potent neutralisation both in vitro and in vivo, offering a potential path to safer, scalable, and globally accessible antivenom therapies. The second example focuses on targeting peptide鈥揗HC complexes that present intracellular cancer antigens on the cell surface. Using generative models, we designed minibinders that specifically recognise these complexes and, when incorporated into immune cell receptors, enable precise killing of tumour cells. Together, these studies demonstrate how AI-guided protein design can move beyond prediction to invention, unlocking new therapeutic strategies across neglected diseases and advanced immunotherapies alike.

Antibodies are coded by nucleotide sequences that are generated by V(D)J recombination and evolve according to nucleotide mutation and selection processes. Existing antibody language models, however, focus exclusively on antibodies as strings of amino acids and are fit using the masked language modeling objective. In this talk, I will first show that fitting using this objective implicitly incorporates nucleotide-level processes as part of the protein language model, which degrades performance when predicting functional properties of antibodies. To address this limitation, we propose a new framework: a deep amino acid selection model (DASM) that predicts the selective effect of replacing every amino acid with every alternate amino acid. By fitting selection as a separate term from the mutation process, the DASM exclusively quantifies functional effects. This separation of concerns leads to substantially improved performance on standard functional benchmarks. Moreover, our model is an order of magnitude smaller and orders of magnitude faster to evaluate than existing approaches, as well as being readily interpretable. I will then describe some surprising conclusions about how natural selection works for antibodies: there is more to the story than framework vs CDRs!

Understanding how the brain dynamically coordinates cognition, emotion, and behavior remains a central challenge in neuroscience and psychiatry. Traditional neuroimaging approaches often rely on static or averaged representations of brain activity, which obscure the rich temporal structure and inter-network transitions that may underlie both healthy and pathological states. In this talk, I will introduce a novel framework that combines topological data analysis (TDA) with graph-theoretic modeling to uncover reproducible and interpretable patterns in brain dynamics. Using the Mapper algorithm, we construct individual-level brain-state graphs that reveal flexible or rigid transitions between functional states across resting-state and task-based fMRI. These dynamic motifs not only reflect traits such as rumination and attentional flexibility but also offer new biomarkers for clinical populations. I will also present our recent efforts to integrate deep learning and latent factor modeling with the Research Domain Criteria (RDoC) framework. This work provides a data-driven refinement of mental health constructs by identifying latent 鈥渕otifs鈥� that better explain individual differences in brain activation across multiple tasks. Together, these approaches highlight the importance of moving beyond summary statistics toward manifold-level insights into brain function, with implications for both precision psychiatry and the development of AI-driven diagnostic tools.

Developmental disorders, including congenital anomalies and neurodevelopmental disorders (NDD), affect many newborns and children. They are major factors associated with mortality, mobility, and long-term health of children and surviving adults. Finding genetic causes is fundamentally important for improving health care of these individuals and advancing basic research in human developmental biology. Previous efforts, which largely focused on de novo variants and loss of function variants with large effect, discovered new risk genes and pathways and were able to identify genetic causes in about 20-40% of 鈥渃omplex鈥� cases who have additional congenital anomalies or NDD. The same approach was much less successful for 鈥渋solated鈥� cases who do not have additional anomalies or NDD. A major analytical gap in previous studies is about rare inherited variants, especially missense variants, due to uncertainty of their genetic consequence and overall low signal-to-noise ratio. We reason that a key to address this issue is to leverage the fact that early onset severe conditions are under strong negative selection, therefore, genetic causes with moderate to large effect are under discernable negative selection. In this talk, I will go through MisFit, a new machine method developed based on population genetics models for estimating selection coefficient of missense variants. I鈥檒l then show new results in autism regarding rare inherited variants and implications for rare congenital anomalies.

Cell state transitions are essential for embryos to develop, wounds to heal and diseases to progress. Still, the mechanisms that coordinate the transitions of cells within tissues from one state to another are not fully understood. In my talk I will discuss our advances in studying how physical forces that emerge from tissue interplay integrate with biochemical frameworks to coordinate robust cell state transitions in time and space.

Biological systems constantly navigate a delicate balance between reproducibility and variability. Developmental processes exemplify remarkable precision鈥攅ach person has five fingers 鈥攂ut neural circuits in the brain must maintain a flexible architecture to enable adaptive behavior. In this talk, I explore this tension through two complementary lenses: small genetic networks guiding precise cell identities in fly development, and large neural networks guiding animal behavior. Specific examples include how electric fish subtract self-generated electrical signals to accurately detect prey, and how zebrafish swim upstream in complete darkness by integrating local water flow rotations in a neural implementation of Stoke鈥檚 theorem. By performing quantitative measurements of gene expression and neural connectivity and relating them to the biological function of each network, I demonstrate that simple rules can emerge even from the most intricate networks.

In this talk I will discuss some recent evidence regarding the degree to which AlphaFold appears to learn to do implicit physics, as well as how this knowledge is acquired during the training process. Time permitting, I will also discuss differences between AlphaFold 2 and 3 with regards to the question of implicit physical knowledge.

Critical points are special places in parameter space where macroscopic system behavior is particularly sensitive to small changes in molecular details. In this talk I will explore two mechanisms through which cells could use a critical point and a dynamical bifurcation, respectively, to sensitively amplify and integrate information arriving at many individually noisy receptors.

First I will discuss the pit organ of certain snakes which forms a low resolution infrared image used for hunting prey. For this organ to be functional, individual neurons innervating the pit must be able to respond to local heating <1mk, at least 1000x more sensitive than the thermo-TRP ion channels which act as molecular receptors. I will present a model for how this remarkable sensitivity can be achieved. In this model, individual channels are electrically coupled, self-tuning to a bifurcation separating a mostly silent 鈥榠rregular spiking鈥� state from an active 鈥榬egular spiking鈥� state. This coupling effectively integrates an order one fraction of the information available in individual channels into the cooperative output of spike frequency.

Second I will discuss our recent work to understand the role of liquid-liquid criticality in signal transduction in eukaryotic cells. This work is motivated by two complementary experimental findings – first, that cellular membranes are two dimensional liquids tuned near a miscibility critical point, and second that many signal cascades begin with the formation of signaling cluster – a membrane domain enriched in particular lipids and membrane bound proteins under which bulk proteins with roles in signaling aggregate. I will argue that these platforms are examples of what we term surface densities; a two-dimensional prewet phase held together by critical casimir forces in the membrane and weak protein-protein interactions that reflect the propensity of many bulk proteins to phase separate. The liquid environment in these surface phases enables certain chemical reactions that are important in the early stages of signaling. As a result, the formation of a prewet phase acts as a crucial step for integrating a signal.

For the last 3.8 billion years, the large-scale structure of evolution has followed a pattern of speciation that can be described by branching trees. Recent work, especially on bacterial sequences, has established that despite their apparent complexity, these so-called phylogenetic or evolutionary trees exhibit two unexplained broad structural features which are consistent across evolutionary time. The first is that phylogenetic trees exhibit scale-invariant topology, which quantifies the fact that their branching lies in between the two extreme cases of balanced binary trees and maximally unbalanced ones. The second is that the backbones of phylogenetic trees exhibit bursts of diversification on all timescales. I present a coarse-grained statistical mechanics model of ecological niche construction coupled to a simple model of speciation, and use renormalization group arguments to show that the statistical scaling properties of the resultant phylogenetic trees recapitulate both the scale-invariant topology and the bursty pattern of diversification in time. These results show in principle how dynamical scaling laws of phylogenetic trees on long time-scales may emerge from generic aspects of the interplay between ecological and evolutionary processes, leading to scale interference.
Finally, I will argue that these sorts of simplistic, minimal arguments might have a place in understanding other large-scale aspects of evolutionary biology. In particular I will mention two questions where we do not have even a qualitative understanding let alone a quantitative one: (1) the spontaneous emergence of the open-ended growth of complexity; (2) the response of evolving systems to perturbations and the implications for their control. Even though biology is intimidatingly complex, “everything has an exception”, and there are a huge number of undetermined parameters, statistical physics reasoning may lead to useful new insights into the existence and universal characteristics of living systems.
Work performed in collaboration with Chi Xue and Zhiru Liu and supported by NASA through co-operative agreement NNA13AA91A through the NASA Astrobiology Institute for Universal Biology.

Bacteria outside of the lab must often navigate complex environments in search of food. A better understanding of how bacteria find their way might help researchers develop strategies to inhibit bacterial infections. To that end, we probe to what extent the common bacteria E. coli explores landscapes that have much in common with structures experienced in nature鈥攆ractals and mazes鈥攁nd find that they appear to have evolved a variety of ways to navigate these puzzles.

Bacteria communicate with each other via chemical signals, which carry information. Loss of information from other bacteria can be viewed as a disturbing event for a bacterium and a signal to avoid that region of space. We used hydrodynamic flow past a small aperture to remove bacterial signaling, with the expectation that bacteria would avoid that aperture due to an effective horizon that absorbed information: a 鈥渂lack hole鈥�. We confirm that a bacterial 鈥淗awking radiation鈥� — collective bacterial waves launching away from the horizon — exists, even in the presence of strong geometrical pumping.

In the final part of the talk, we present an ecology-inspired form of active matter consisting of a robot swarm on an interactive light-emitting diode lightboard.

The menstrual cycle is a powerful indicator of overall health in women. However, its complete characterization remains an open research question, in part due to a lack of direct and reliable measurements of menstruation over-time and across individuals. In this talk, I will describe how statistical learning techniques can provide a better understanding of the menstrual cycle and its patterns, by accommodating different types of (noisy and missing not at random) observations: reproductive hormone level measurements through time and self-reported cycle start dates from menstrual trackers. The former are invasive and expensive to attain, but provide a detailed, in-depth view; the latter are easy and cheap to attain, but shallow and subject to biases. I will present statistical learning solutions that accommodate these noisy physiologic and self-tracked variables, towards robust, personalized models that can characterize menstruation.

Electrical signaling in biology is typically associated with action potentials, transient spikes in membrane voltage that return to baseline. More generally, electrical tissues are reaction-diffusion systems which could support patterns which are structured in space but stationary in time. It is possible that electrical signaling could coordinate biological processes on timescales which are slower than action potentials 鈥� for example, during embryonic development. Constraints on electrophysiological measurement in vivo have made it challenging to assess these hypotheses quantitatively.
Here we present a new strategy to study biological pattern formation by building synthetic bioelectrical tissues from the bottom-up. We engineer electrically inert mammalian cells to express ion channels, optogenetic actuators, and fluorescent voltage indicators in tandem. By combining patterned illumination and all-optical electrophysiology, we study these synthetic bioelectric tissues as excitable media and compare their dynamics rigorously to mathematical predictions.
In one demonstration, we engineer bioelectrical circuits of spiking cells capable of simple information processing and memory(1) We also show that dynamical transitions between stable and arrhythmic spiking patterns in these tissues depends sensitively on tissue geometry and dimensionality(2). Finally, we show that synthetic tissues of electrically bistable cells can form spatially structured but time-stationary electrical domains which polarize via nucleation-and-growth phase transitions(3). Observation of electrical domains in a stem cell model of myogenesis suggest that bioelectrical phase transitions may play a physiological role during embryonic development.

1. McNamara, H.M., Zhang, H., Werley, C.A. and Cohen, A.E., 2016. Optically controlled oscillators in an engineered bioelectric tissue. Physical Review X, 6(3), p.031001.
2. McNamara, H.M., Dodson, S., Huang, Y.L., Miller, E.W., Sandstede, B. and Cohen, A.E., 2018. Geometry-dependent arrhythmias in electrically excitable tissues. Cell Systems, 7(4), pp.359-370.
3. McNamara, H.M., Salegame, R., Al Tanoury, Z., Xu, H., Begum, S., Ortiz, G., Pourquie, O. and Cohen, A.E., 2019. Bioelectrical signaling via domain wall migration. bioRxiv, p.570440. (in press, Nature Physics).

This talk addresses some statistical and computational problems arising in the study of cancer evolution. The starting point comes from population genetics: how should we estimate evolutionarily relevant parameters from DNA sequence data taken from samples of individuals?
I will give a brief overview of what we learned, touching on Approximate Bayesian Computation as an inference method when likelihoods are intractable. To illustrate ABC I will give an example concerning inference about the number of distinct DNA sequences in a sample, given only information about the relative frequency of point mutations in the samples. This provides an introduction to inference from typical cancer sequencing data, in which individuals are replaced by cells and in which typically we do not know which mutations occur in which cells. I will discuss a stochastic model that exploits coalescent theory to study clonal sweeps, and describe new techniques for deconvolving clones from single cell sequencing data. Time permitting, I will describe some novel experimental methods we are developing to understand the 3D structure of tumors, paving the way for some challenging inferential problems that will require engagement from data scientists and others.

Microvascular networks are complex networks containing up to tens of thousands of blood vessels in a millimeter cube. While the main function of these networks is transporting oxygen and nutrients to tissues, there are other functions important to survival, such as transport efficiency and damage resistance, and it is not clear which functions are adaptive for microvascular networks. Here we start by looking at zebrafish trunk microvascular network. The fine vessels are like rungs of a ladder with different distances from the heart, yet the blood flows through them do not exponentially decrease as predicted by simple electric circuit approximation. The red blood cell temporarily occludes the fine vessel it passes through, and we found that this occlusion is tuned for achieving uniform blood flow in fine vessels. The uniform flow is maintained at the cost of transport efficiency, suggesting that microvascular networks might not be transport efficient. To further explore this hypothesis, we implemented a gradient descent algorithm that finds optimal networks for general functions and constraints, and we compared zebrafish trunk network with uniform flow network and transport efficient network. We found that the uniform flow network quantitatively describes the real zebrafish trunk network. When transport efficiency is imposed as a constraint instead of a target function, we discovered an abrupt change in network morphology when the burden of transport cost exceeds a threshold, suggesting that biological network always conforms to the most urgent constraint on it. Finally, I will talk about our recent work on a novel vessel adaptation mechanism based on sensing the large shear stress created by red blood cells, and compare our prediction to blood flow data in zebrafish embryos of different ages.

All morphogenesis processes, whether at the cell scale or tissue level, require the coordination of growth and mechanics to properly shape functional structures. However, the mechanisms that coordinate these two processes in the sculpting of individual cells, and especially in walled cells, remain unknown. Using yeast mating projection growth as a model system, we show that a genetically-encoded mechanical feedback relays information about the mechanical state of the cell wall to the intracellular processes assembling it, thereby coordinating cell wall expansion and growth during cell morphogenesis. We find that mechanical feedback is essential to stabilize cell growth, but the shape and size of the cell are insensitive to the feedback and independently controlled. Beyond the role of mechanical feedback in cell viability during growth, cell wall mechanics may also inform other key cellular processes, such as cell polarization. While simulations of the cell polarization machinery can readily generate a stable polarization cap in a spherical geometry, maintenance of the polarization cap at the tip of the mating projection is not possible in realistic cell geometries. We show that the same mechanical feedback mechanism ensuring cell viability also results in the desired spatial coordination of the location of the polarization cap and the tip of the projection.

How is the physical state of DNA maintained so that the right genes are activated, controlled and silenced at the right time and in the right place? How do cells with the same genome di fferienate into cell types which display such di ferent behavior? How are those behaviors maintained? And how do they go wrong? Answering these questions has emerged as a grand challenge for molecular biology in the coming decade. In this talk, I will outline some of the basic physical elements of transcriptional initiation and elongation. A mathematical formulation of the ‘twin domain model’ of transcription will be presented with a focus on simple descriptions of each process. The resulting framework is combined with well-established stochastic processes of transcription resulting in a model which characterizes the impact of the mechanical properties of transcription on elements of gene expression and DNA structure. Importantly, this model opens a window onto the role of gene expression in shaping the storage and transfer of information through the physical and chemical state of DNA offering insight into a mechanical method of feedback for cells to regulate function.

Tens of thousands of biochemical reactions occur simultaneously in the cell. Small molecules are channeled through metabolic pathways at blistering speed. Giant complexes assemble to orchestrate transcription and translation. ATP fuels the active transport of organelles along microtubules, and actin networks drive membrane remodeling and agitate the cytoplasm. All of this occurs within a crowded cell interior that approaches the physical limits where jamming will occur. This extreme physical environment is both essential for life, and a potential liability. If cells become too dilute, they senesce and die. On the other hand, increased crowding eventually stalls growth. We found that the central growth regulator, mTORC1, is a crucial regulator of crowding. We also found that mechanical compression affects crowding and phase separation. We propose that perturbations to the physical properties of the cell interior through mechanical forces and the regulation of growth pathways play a crucial role in cell, developmental and disease biology, including cancer and neurodegeneration.

In this talk I will present work done in my lab showing how ecological interactions control the assembly and function of microbial communities at micro-scales. Using model marine particles composed of a variety of carbohydrate commonly found in the ocean, I will show how microbial interactions such as cross-feeding and social cheating control community dynamics, leading to rapid successions on particles. By comparing successions on different substrates we were able to cluster successional species into two types of functional groups, one type that is specific to each carbohydrate, and a type that substrate-unspecific and instead driven by crossfeeding interactions. We show that this simple logic can be exploited to predict the composition of communities on multiple carbohydrates as simple linear combinations of the compositions on single carbohydrates. Finally, I will focus on the substrate-specific group of chitin degraders to discuss ongoing work on the role of cooperative interactions, mediated by extracellular enzymes.

Unique physical, chemical, and optical phenomena arise when materials are confined to the nano-scale. We are accustomed to making observations and predictions for the behavior of living systems on a macroscopic scale that is intuitive for the time and size scales of our day-to-day lives. However, the building blocks of life: proteins, nucleic acids, and cells, occupy different spatiotemporal scales. Our lab focuses on understanding and exploiting tunable optical and chemical properties of nanomaterials to access information about biological systems stored at the nano-scale. We present recent work on developing and implementing dopamine nanosensors to image dopamine volume transmission in the extracellular space of the brain striatum in both model (mouse) and non-model (bat) species. We validate our dopamine nanosensor in acute striatal slices with electrical and optogenetic stimulation of dopamine release, and show disrupted dopamine release or reuptake kinetics when brain tissue is exposed to dopamine agonist or antagonist drugs [1]. We characterize our findings in the context of their utility for high spatial and temporal neuromodulator imaging in the brain with 2-photon microscopy [2], and describe nanosensor exciton behavior from a molecular dynamics (MD) perspective [3]. We also discuss how high aspect ratio nanomaterials can be synthesized to carry biomolecular cargo to living systems. In particular, genetic engineering of plants is at the core of environmental sustainability efforts, but the physical barrier presented by the cell wall has limited the ease and throughput with which exogenous biomolecules can be delivered to plants. We will describe how nanomaterials engineering principles can be leveraged to manipulate living plants, in efforts to reconcile the benefits of crop genetic engineering with the demand for non-GMO foods [4]. Our work in the agricultural space provides a promising tool for species-independent, targeted, and passive delivery of genetic material, without transgene integration, into plant cells for rapid and parallelizable testing of plant genotype-phenotype relationships.

1. Beyene, A.B., McFarlane, I.R., Pinals, R.L, Landry, M.P.鈥� Stochastic Simulation of Dopamine Neuromodulation for Implementation of Fluorescent Neurochemical Probes in the Striatal Extracellular Space. ACS Chemical Neuroscience 8 (10), 2275-2289 (2017).
2. Del Bonis O鈥橠onnell, J.T., Page, R.H., Beyene, A.G., Tindall, E.G., McFarlane, I.R., Landry, M.P.鈥� Molecular Recognition of Dopamine with Dual Near Infrared Excitation-Emission Two-Photon Microscopy. Advanced Functional Materials (2017). DOI: 10.1002/adfm.201702112
3. Beyene, A.G., Delevich, K., Del Bonis-O鈥橠onnell, J.T., Piekarski, D.J., Lin, W.C., Thomas, A.W., Yang, S.J., Kosillo, P., Yang, D., Wilbrecht, L., Landry, M.P. 鈥� Imaging Striatal Dopamine Release Using a Non-Genetically Encoded Near-Infrared Fluorescent Catecholamine Nanosensor. Nano Letters 18 (11), 6995-7003 (2018).
4. Demirer, G.S., Chang, R., Zhang, H., Chio, L., Landry, M.P.鈥� Nanoparticle-Guided Biomolecule Delivery for Transgene Expression and Gene Silencing in Mature Plants. bioRxiv (2018). DOI: 10.1101/179549

Motile cilia are micron-scale hair-like protrusions from epithelial cells that beat collectively to transport fluid. On the tissue level, cilia serve diverse biological functions, such as mucociliary clearance in the airways and cerebrospinal fluid transport in the brain ventricles. Yet, the relationship between the structure and organization of ciliated tissues and their biological function remains elusive. Here, I will present a series of mathematical and experimental models that examine: (1) the emergence of self-sustained oscillations in a single cilium, (2) the coordinated beating of neighboring cilia, and (3) the role of cilia-driven flows in particle transport, mixing, capture and filtering. I will conclude by commenting on the implications of these models to understanding the biophysical mechanisms underlying the interaction of ciliated tissues with microbial partners.

Computational recognition of distant sequence homology is a key to studying ancient events in molecular evolution. The better our sequence analysis methods are, the deeper in evolutionary time we can see. A major aim in the field is to improve the resolution of homology recognition methods by building increasingly realistic, complex, parameter-rich models. I will describe current and future research in homology search algorithms based on probabilistic inference methods, including hidden Markov models (HMMs) and stochastic context-free grammars (SCFGs). We make these methods available in the HMMER and Infernal software from my laboratory, in collaboration with sequence family databases including Pfam and Rfam.

Phenotypic diversity and collective behavior are important properties of living communities that provide selective advantages. While collective behavior requires coordination between individuals, phenotypic diversity tends to reduce coordination. I will report on our recent experimental and theoretical results that used bacterial chemotaxis as model system to examine how phenotypic heterogeneity modulates group performance and how collective behavior affects the amount of diversity in the group.

This study was supported by the National Institutes of Health grant R01GM106189, the Allen Distinguished Investigator Program (grant 11562) through The Paul G. Allen Frontiers Group, and the James S. McDonnell Foundation grant on Complexity.

I will briefly describe scalable models for regulatory network inference that enable multiple study and multiple cell-type comparisons, include some of the key biophysics of transcriptional dynamics, and can be adapted to the latest single-cell and spatial genomics technologies. I will focus on new developments to enable learning networks from single cell and spatial transcriptomics. A key challenge in single-cell and spatial transcriptomics data is that many heterogenous cell types are measured in each experiment; to address this heterogeneity we adapt multi-task approaches to network inference to the context of single cell data. Another key challenge with spatial and single-cell based network inference is the lack of appropriate benchmarks. To address this lack of a suitable benchmark dataset, we (with David Gresham lab) have generated a new single-cell RNA sequencing data set for the model species Saccharomyces cerevisiae (there are many thousands of well characterized regulatory and signaling interactions for this well studied model system), which contains 40,000 individual cells (cells grown in 11 different environmental growth conditions and on multiple genetic backgrounds; TF knockouts). These new scRNA-seq datasets are integrated with ATAC-seq and other genomic resources to adapt and benchmark our network inference tools. We will also provide this dataset to the computational community for tool development any day now. These methods are currently being applied to work on the mammalian immune system, cancer, mouse intercortical neuron development and the spatiotemporal genomics of human and mouse spinal column.

The single cell biflagellate Chlamydomonas reinhardtii has proven to be a very useful model organism for studies of size control. The lengths of its two flagella are tightly regulated. We study a model of flagellar length control whose key assumption is that proteins responsible for the intraflagellar transport (IFT) of tubulin are present in limiting amounts. In the case of two simultaneously assembling flagella, regardless of the details of how the flagella are coupled, we find that the widely-used assumption of a constant disassembly rate is inconsistent with experimental results. We therefore propose a model in which diffusion gives rise to a length-dependent concentration of depolymerizer at the flagellar tip. This model is found to be consistent with experimental results and generalizes to other situations such as arbitrary flagellar number.

One of the grand goals in science is to understand how neural, genetic, and biochemical circuits produce behavior. While a great deal of work has been done in the development of tools to perturb and measure the circuits underlying sensory behavior, advances in the study of behavior itself has lagged behind. For this talk I will describe some attempts to close this gap with focus on C. elegans locomotion and its response to thermal stimuli. The roundworm C. elegans is a simple organism with only 300 neurons but it can generate complex adaptive behavioral responses to a wide range of sensations including taste, touch, and temperature. C. elegans behavior consists of a number of stereotyped locomotory states such as forward and reverse, pausing, turning, etc. Typically, the worm moves by making stochastic transitions between these states, and in response to sensory measurements it adapts by biasing the probability of these transitions. However, under certain conditions the worm will transition from stochastic to deterministic behavior such as in response to sensory gradients or noxious or 鈥減ainful鈥� stimuli. We have developed simple desktop experiments to quantitatively capture the behavioral responses of C. elegans to precise thermal stimuli. Using these data we have shown that C. elegans moves through a 鈥渟hape space鈥� that is low dimensional in which four dimensions capture approximately 95% of the variance in body shape. Here I will give two examples of modeling that take advantage of this low dimensionality and stereotypy. In the first we show that stochastic dynamics within this shape space predicts transitions between attractors corresponding to abrupt reversals in crawling direction. With no free parameters, our inferred stochastic dynamical system generates reversal timescales and stereotyped trajectories in close agreement with experimental observations. In the second we use Sir Isaac, an algorithm that allows inference of the dynamical equations underlying a noisy time series, even if the dynamics are nonlinear鈥攖o analyze the thermal 鈥減ain鈥� response of C. elegans. Both examples show that it is possible to learn 鈥渆quations of behavior鈥� of the worm, and that these equations give an interpretable, complementary perspective to traditional biological studies.

The ability of the brain to encode and store information depends on the plastic nature of the individual synapses. The increase and decrease in synaptic strength, mediated through the structural plasticity of the spine, are important for learning, memory, and cognitive function. Dendritic spines are small structures that contain the synapse. They come in a variety of shapes (stubby, thin, or mushroom-shaped) and a wide range of sizes that protrude from the dendrite. These spines are the regions where the postsynaptic biochemical machinery responds to the neurotransmitters. Spines are dynamic structures, changing in size, shape, and number during development and aging. While spines and synapses have inspired neuromorphic engineering, the biophysical events underlying synaptic and structural plasticity remain poorly understood.

Our current focus is on understanding the biophysical events underlying structural plasticity. I will discuss two recent efforts from my group 鈥� first, a systems biology approach to construct a mathematical model of biochemical signaling and actin-mediated transient spine expansion in response to calcium influx caused by NMDA receptor activation and second, a series of spatial models to study the role of spine geometry and organelle location within the spine for calcium and cyclic AMP signaling. I will conclude with some new efforts in using reconstructions from electron microscopy to inform computational domains. I will conclude with how geometry and mechanics plays an important role in our understanding of fundamental biological phenomena and some general ideas on bio-inspired engineering.

There has long been controversy and conflicting data regarding the speed at which flagellated bacteria swim in fluids of different Newtonian and non-Newtonian properties. Some data indicates faster swimming in viscoelastic media, while others suggest slower swimming. There is even controversy regarding the swimming speed in purely Newtonian media of different viscosities. We will present data and analysis from recent experiments in which we follow smooth-swimming and wild-type E coli using a tracking microscope as they move through different Newtonian and non-Newtonian fluids. We find that increasing the bulk viscosity affects swimming speed in two ways – it reduces the effectiveness of flagellar propulsion, but also increases the flagellar bundling time of wild type E coli. Cells are found to swim faster in non-Newtonian fluids, even though the bulk viscosity rises. The dominate contribution is found to be related to the strong shear thinning property of the fluid, although the normal stresses associated with viscoelasticity can also contribute through reduced cell precession, slightly improved propulsive efficacy and faster flagellar bundling times.

Studies of signaling cascades can reveal important mechanisms driving multicellular development, but the models that emerge often lack critical links to environmental cues and metabolites. We study the effects of extra- and intracellular chemistry on biofilm morphogenesis in the pathogenic bacterium Pseudomonas aeruginosa, which produces oxidizing pigments called phenazines. While wild-type colonies are relatively smooth, phenazine-null mutant colonies are wrinkled. Initiation of wrinkling coincides with a maximally reduced intracellular redox state, suggesting that wrinkling is a mechanism for coping with electron acceptor limitation. Mutational analyses and in situ expression profiling have revealed roles for PAS-domain and other redox-sensing regulatory proteins, as well as genes involved in motility and matrix production, in colony morphogenesis. To characterize endogenous electron acceptor production, we have developed a novel chip that serves as a growth support for biofilms and allows electrochemical detection and spatiotemporal resolution of phenazine production in situ. Through these diverse approaches, we are developing a broad picture of the mechanisms and metabolites that exert an integrated influence over redox homeostasis in P. aeruginosa biofilms

Cell-free expression has crossed the ages as a tool to address fundamental questions in biology and as an experimental platform to prototype biochemical systems. In the last decade, a new post-genomic generation of cell-free expression systems has emerged. In vitro protein synthesis has become powerful, versatile and scalable. Larger and larger DNA programs can be executed in vitro to construct living systems from scratch. Cell-free expression has also been devised to expand the capabilities of natural cells. In this talk, I will present the current field of cell-free constructive biology and show several examples of current capabilities, including prototyping CRISPR technologies and assembling synthetic cells.

Artificial cells made of molecular components and lipid membrane are emerging platforms to characterize, by construction, the properties of living systems. Cell-free transcription-translation (TXTL) system offers several advantages for the bottom-up synthesis of cellular reactors. Yet, scaling up their design within well-defined geometries remains challenging. We present a microfluidic device hosting TXTL reactions of a single reporter gene in thousands of microwells separated from an external buffer by a phospholipid membrane. The yield of the protein synthesis in sealed micro-reactors exhibits a bimodal distribution of active and inactive states, in spite of the absence of transcriptional feedback regulation. Adding fatty acid and polymer in buffer increases the membrane stability along with the proportion of active micro-reactors. Protein synthesis is greater in micro-reactors with stable membranes. This indicates that boundary plays regulatory role on cell-free TXTL systems as a part of minimal artificial cells.

Synonymous codon choice can have dramatic effects on ribosome speed, RNA stability, and protein expression. Ribosome profiling experiments have underscored that ribosomes do not move uniformly along mRNAs, exposing a need for models of translation that capture the full range of empirically observed variation. Previously, we showed that deep sequencing of ribosome-protected mRNA fragments reveals not only the position of each ribosome but also, unexpectedly, its particular stage of the elongation cycle. This provides a new way to study the detailed kinetics of translation and a new probe with which to identify sequence elements that affect each step in the elongation cycle. More recently, we modeled variation in translation elongation using a feedforward neural network to predict the ribosome density at each codon as a function of its sequence neighborhood. We applied our model to design synonymous variants of a fluorescent protein in yeast and concluded that control of translation elongation alone is sufficient to produce large, quantitative differences in protein output.

The behavior of an organism often reflects a strategy for coping with its environment. Such behavior in higher organisms can often be reduced to a few stereotyped modes of movement due to physiological limitations, but finding such modes in amoeboid cells is more difficult as they lack these constraints. In this talk, I will examine cell shape and movement in starved Dictyostelium amoebae during single-cell and collective cell migration. I will show that the incredible variety in amoeboid shape across a population can be reduced to a few modes of variation, and that different modes are used depending on the steepness of the applied chemical gradient or drug treatment. The origins of cell shape and behavior are tackled by both forward and inverse modeling, predicting long-term cell behavior without reference to biochemical details. Analogies with statistical mechanics are made throughout the talk.

The classical picture of a neural network assumes that each neuron sums its inputs, followed by a nonlinear activation function. Interesting computations emerge by combining and wiring up these individual units. While immensely successful (both in neuroscience and AI), this view has also created several persistent puzzles about the organization of neural systems. One puzzle are spikes, which have largely remained a nuisance, rather than a feature of neural systems. Another puzzle is robustness to perturbations, which is ubiquitous in biology, but largely ignored in neural network modeling. I will show that these puzzles can be resolved if we shift our perspective of how neural systems operate. Based on only two assumptions – that the effective output of a neural network can be extracted via linear readouts, and that each neuron only fires to bound an error on this output, I will show how to derive a spiking network of integrate-and-fire neurons that exhibits irregular and asynchronous spike trains, balance of excitatory and inhibitory currents, and robustness to perturbations. I will provide geometric intuitions for the network’s functionality, and use these insights to show how these networks can control the precision of their output via bottom-up and top-down gain modulation, even though they consist of integrate-and-fire neurons only.

There is an enormous potential for evolution within each of our microbiomes, with billions of new mutations being created each day. In this talk, I will highlight the power of tracking within-person evolution for understanding bacterial transmission, identifying genes critical to long-term survival, and for understanding evolutionary principles. I will present examples from infectious diseases, the gut microbiome, and the skin microbiome, and speculate on how within-person evolution may impact and inform microbiome-targeted therapies.
Relevant DOIs: ,

Amyloid-forming peptides or proteins typically form amyloid fibrils through a nucleation and growth mechanism. The nucleation step typically requires a minimum concentration of peptides in the solution, thus preventing this one-dimensional crystallization process from moving forward in dilute solutions. We demonstrate that in a dilute solution, where bulk nucleation and growth is limited, providing the right surface can promote rapid self-assembly of peptides into mono-layer thick amyloid fibrils within minutes. This process is diffusion-limited and depends on the orientation and diffusion of peptides on the surface. The deposition rate of peptides, as well as their orientation and diffusion on the surface strongly depend on the strength of the peptide-surface interactions. The self-assembly can be suppressed either by reducing the interaction such that the deposition rate is reduced or by increasing the interaction energy to a point where the peptide diffusion is substantially reduced.
Once the self-assembly moves forward with time, the side-chains of the residues can modulate the surface energy and thus the interaction energy between the incoming peptides and the effective surface. As such, depending on the nature of the side chains, the surface can become more hydrophobic or hydrophilic and promote or prevent the further deposition of the peptides. We demonstrate the generality of using this method to modulate the effective surface energy by the desired amount through the design of side chains, thus engineering surfaces with self-limiting self-assemblies with specific surface energy values.

Deciphering chromosome tertiary organization is essential for understanding how genetic information is replicated, transcribed, silenced, and edited to control basic life processes. Many experimental studies of chromatin using nucleosome structure determination, ultra-structural techniques, single-force extension studies, and analysis of chromosomal interactions have revealed important chromatin characteristics as a function of various internal and external conditions, such as looping, compaction, and compartmentalization. Modeling studies, anchored to high-resolution nucleosome models, have explored related questions systematically. In this talk, I will describe multiscale computational approaches for chromatin modeling at nucleosome resolution and recent mesoscale chromatin simulations that incorporate key physical parameters such as nucleosome positions, linker histone binding, and acetylation marks to ‘fold’ in silico the Hox C gene cluster. The folded gene reveals a contact hub that connects an acetylation-rich with a linker histone-rich region. Such chromatin modeling techniques open the way to other computational folding of genes and genomes. Moreover, the resulting folded system emphasizes the heterogeneity of chromatin fibers and hierarchical looping motifs, and underscores how nucleosome positions in combination with epigenetic marks and linker histone binding direct the tertiary folding of fibers and genes to perform their cellular tasks. These chromatin architecture findings have important implications on many important processes including cell differentiation, gene regulation, and disease progression.
Of possible interest:
G. Ozer, A. Luque and T. Schlick, “The Chromatin Fiber: Multiscale Problems and Approaches”, Curr. Opin. Struc. Biol. 31: 124–139 (2015).
S. Grigoryev, G. Bascom, J. M. Buckwalter, M. Schubert, C. L. Woodcock, and T. Schlick, “Hierarchical Looping of Zigzag Nucleosome Chains in Metaphase Chromosomes”, Proc. Natl. Acad. Sci. USA 113: 1238–1243 (2016).
G. Bascom, K. Sanbonmatsu, and T. Schlick, “Mesoscale Modeling Reveals Hierarchical Looping of Chromatin Near Gene Regulatory Elements”, J. Phys. Chem. B Special Issue: J. Andrew McCammon Festschrift 120: 8642–8653 (2016).
G. Bascom and T. Schlick, “Linking Chromatin Fibers to Gene Folding by Hierarchical Looping”, Biophys. J. 112: 434鈥�445 (2017).
G. Bascom, T. Kim, and T. Schlick, “Kilobase Pair Chromatin Fiber Contacts Promoted by Living-System-Like DNA Linker Length Distributions and Nucleosome Depletion”, J. Phys. Chem. B Special Issue in Memory of Klaus Schulten 121 (15): 3882–3894 (2017).
S. S. P. Rao, S.-C Huang, B. G. St. Hilaire, J. M. Engreitz, E. M. Perez, K.-R Kieffer-Kwon, A. L. Sanborn, S. E. Johnstone, G. D. Bascom, I. D. Bochkov, X. Huang, M. S. Shamim, J. Shin, D. Turner, A. D. Omer, J. T. Robinson, T. Schlick, B. E. Bernstein, R. Casellas, E. Lander, and E. Lieberman-Aiden, “Cohesin Loss Eliminates All Loop Domains, Leading to Links Among Superenhancers and Downregulation of Nearby Genes”, Cell 171: 305–320 (2017).
G. Bascom, C. Myers, and T. Schlick, “Mesoscale Modeling Reveals Formation of an Epigenetically Driven HOXC Gene Hub”, Proc. Natl. Acad. Sci. USA 116: 4955–4962 (2019). (doi: https://doi.org/10.1073/pnas.1816424116). Accompanying commentary by Michele D. Pierro, “Inner Workings of Gene Folding”, Proc. Natl. Acad. Sci. USA 116: 4774–4775 (2019).
O. Perisic, S. Portillo-Ledesma, and T. Schlick, “Sensitive Effect of Linker Histone Binding Mode and Subtype on Chromatin Condensation”, Nuc. Acids Res., doi: 10.1093/nar/gkz234 (2019).

Human behavior is founded on the ability to identify meaningful entities in complex noisy data streams that constantly bombard the senses. For example, in vision, retinal input is transformed into rich object-based scenes; in audition, sound waves are transformed into words and sentences. In this talk, I will describe my work using computational models to help uncover how sensory cortex accomplishes these enormous computational feats.
The core observation underlying my work is that optimizing neural networks to solve challenging real-world artificial intelligence (AI) tasks can yield predictive models of the cortical neurons that support these tasks. I will first describe how we leveraged recent advances in AI to train a neural network that approaches human-level performance on a challenging visual object recognition task. Critically, even though this network was not explicitly fit to neural data, it is nonetheless predictive of neural response patterns of neurons in multiple areas of the visual pathway, including higher cortical areas that have long resisted modeling attempts. Intriguingly, an analogous approach turns out be helpful for studying audition, where we recently found that neural networks optimized for word recognition and speaker identification tasks naturally predict responses in human auditory cortex to a wide spectrum of natural sound stimuli, and help differentiate poorly understood non-primary auditory cortical regions. Together, these findings suggest the beginnings of a general approach to understanding sensory processing the brain.
I’ll give an overview of these results, explain how they fit into the historical trajectory of AI and computational neuroscience, and discuss future questions of great interest that may benefit from a similar approach.

Living cells generate and transmit mechanical forces over diverse time-scales and length-scales to determine the dynamics of cell and tissue shape during both homeostatic and pathological processes, from early embryonic development to cancer metastasis. These forces arise from the cell cytoskeleton, a scaffolding network of entangled protein polymers driven out-of-equilibrium by enzymes that convert chemical energy into mechanical work. However, how molecular interactions within the cytoskeleton lead to the accumulation of mechanical stresses that determine the dynamics of cell shape is unknown. Furthermore, how cellular interactions are subsequently modulated to determine the shape of the tissue is also unclear. To bridge these scales, our group in collaboration with others, uses a combination of experimental, computational and theoretical approaches. On the molecular scale, we use active nematic liquid crystals as a framework to understand how mechanical stresses are produced and transmitted within the cell cytoskeleton. On the scale of cells and tissues, we abstract these stresses to surface tension in a liquid droplet and draw analogies between the dynamics of droplet wetting (and dewetting) and the shape dynamics of cells and simple tissues. Together, we attempt to develop comprehensive description for how cytoskeletal stresses translate to the physical behaviors of cells and tissues with significant phenotypic outcomes such as cancer metastasis.

The brain builds its internal sensory representations based on the structure of neural activity. A number of modalities, such as the early visual system and the spatial map in hippocampus, are relatively well-characterized. However some sensory systems, such as olfaction, remain enigmatic. Perhaps the primary difficulty is that the underlying perceptual space is not well-understood. Can we “build” the sensory space from neural activity alone, without a prior understanding of how the stimuli are organized? I will describe a set of mathematical tools that allow to infer the dimension and the topology of stimulus space and illustrate their utility for two neural systems: hippocampus and early olfaction.

Blood is a suspension of objects of various shapes, sizes and mechanical properties, whose distribution during flow is important in many contexts. Red blood cells tend to migrate toward the center of a blood vessel, leaving a cell-free layer at the vessel wall, while white blood cells and platelets are preferentially found near the walls, a phenomenon called margination that is critical for the physiological responses of inflammation and hemostasis. Additionally, drug delivery particles in the bloodstream also undergo margination 鈥� the influence of these phenomena on the efficacy of such particles is unknown.

In this talk a mechanistic theory is developed to describe segregation in blood and other confined multicomponent suspensions. It incorporates the two key phenomena arising in these systems at low Reynolds number: hydrodynamic pair collisions and wall-induced migration. The theory predicts that the cell-free layer thickness follows a master curve relating it in a specific way to confinement ratio and volume fraction. Results from experiments and detailed simulations with different parameters (flexibility of different components in the suspension, viscosity ratio, confinement, among others) collapse onto the same curve. In simple shear flow, several regimes of segregation arise, depending on the value of a “margination parameter” M. Most importantly, there is a critical value of M below which a sharp “drainage transition” occurs: one component is completely depleted from the bulk flow to the vicinity of the walls. Direct simulations also exhibit this transition as the size or flexibility ratio of the components changes. Results are presented for both Couette and plane Poiseuille flow. Experiments performed in the laboratory of Wilbur Lam indicate the physiological and clinical importance of these observations.

Microscopically diverse systems often yield to surprisingly simple effective theories. The renormalization group (RG) describes how system parameters change as the scale of observation grows and gives a precise explanation for this emergent simplicity in physics. We use information geometry to reformulate the RG as a statement about how the distinguishability of microscopic parameters depends on the scale of observation. We show that information about relevant parameters is preserved, with distances along relevant directions maintained under flow. By contrast, irrelevant parameters become less distinguishable under the flow, with distances along irrelevant directions contracting. We apply our tools to understand the emergence of the diffusion equation and more general statistical systems described by a free energy. This suggests a way to identify relevant directions for more general coarsening procedures.

Adaptation is a ubiquitous trait of living organisms in dealing with the outside world. At the cellular level, high sensitivity over a broad range of environmental conditions is achieved via biomolecular circuits that operate out of thermal equilibrium. Here we report an unexpected link between sensory adaptation and auto-induced collective oscillations in dense cell populations. We uncover a frequency regime where adaptive cells amplify temporal variations of the extracellular signal. Deeply rooted in nonequilibrium thermodynamics, the design principle unites several known examples of dynamical quorum sensing, and provides a new perspective on regulatory mechanisms behind glycolytic oscillations in yeast cell suspensions.

Mechano-chemical processes in thin biological structures, such as the cellular cortex or epithelial sheets, play a key role during the morphogenesis of cells and tissues. Emergent dynamics in these processes can arise from a feedback loop in which active mechanical forces, by inducing material flows, indirectly affect their own chemical regulation. It has been demonstrated in simple, fixed geometries that this mechanism enables self-organized patterning, but the interplay of mechano-chemical processes with complex surface geometries and shape changes of the material remains to be explored. In this work, we employ the theory of active surfaces and develop analytical and numerical tools to study these materials in curved and dynamically evolving geometries. Additionally, diffusive and advective transport processes can redistribute molecules responsible for local stress generation within the surface, which resembles the interplay between active forces, the shape changes they imply and the effects this has on their regulation. Within this framework, cell polarization, contractile ring formation before cytokinesis, as well as pulsatile dynamics in active tubular structures can be understood as natural emergent phenomena. Our approach provides novel opportunities to explore different scenarios of mechano-chemical self-organization and can help understand the role of shape as an effective regulatory element in morphogenetic processes.

Recent advances in neural network modelling have enabled major strides in computer vision and other artificial intelligence applications. This brain-inspired technology provides the basis for tomorrow鈥檚 computational neuroscience [1]. Deep convolutional neural nets trained for visual object recognition have internal representational spaces remarkably similar to those of the human and monkey ventral visual pathway [2]. Functional imaging and invasive neuronal recording provide increasingly rich measurements of brain activity in humans and animals, but a challenge is to leverage such data to gain insight into the brain鈥檚 computational mechanisms [3, 4].

In my lab, we build neural network models of primate vision, inspired by biology and guided by engineering considerations [5]. We also develop statistical inference techniques that enable us to adjudicate between complex brain-computational models on the basis of brain and behavioral data [3, 4]. I will discuss recent work extending deep convolutional feedforward vision models by adding recurrent signal flow and stochasticity, two characteristics of biological neural networks. This improves inferential performance and enables neural networks to more accurately represent their own uncertainty.

[1] Kriegeskorte N (2015) Annu. Rev. Vis. Sci. 2015. 1:417-46.

[2] Khaligh-Razavi SM, N Kriegeskorte (2014) PLoS Computational Biology

[3] Diedrichsen J, Kriegeskorte N (2017) PLoS Computational Biology

[4] Kriegeskorte N, Diedrichsen (2016) J Phil. Trans. R. Soc. B.

[5] Spoerer CJ, McClure P, Kriegeskorte N (2017) Frontiers.

I will present two projects demonstrating the interplay between growth and dispersal in evolutionary contexts. In the first one, I focus on the evolution of expanding populations, such as bacterial colonies or cancerous tumors. Population expansion is driven by growth and dispersal.
The evolutionary fate of a mutant also depends on both growth and dispersal. For instance, when could moving faster and growing slower be a favorable strategy? Starting with the Fisher-KPP equation, I come up with a quantitative rule of invasion, which depends on growth and dispersal rates of the ancestor and the mutant. Theoretical findings are supported by experiments with bacterial swarming colonies, as well as data from ecology literature.
In the second project, I address the question of design optimization of a biochemical network (the C-di-GMP network in bacteria) that integrates signals from the environment and regulates genes for growth and dispersal. This network needs to be trained to provide the optimal outcome when the environment alternatively favors growth or dispersal. I show that natural selection in this network architecture is mathematically equivalent to training a machine learning model to solve a classification problem.

Developmental biology is at a special point in its history: 1) Model organisms have been extensively characterized, 2) the molecular parts list has largely been enumerated, and 3) live-imaging and sequencing tools have matured. What remains is to understand how the system works. I will present work in progress in collaboration with Richard Carthew that focuses on a diverse set of phenomena made manifest during Drosophila eye development. Incomplete results related to collective cell-fate determination, collective polarization, and mechanics will be presented. An emphasis will be placed on the importance of taking a dynamical view of development, and mathematical modeling as a way to synthesize disparate observations, guide experimentation, and make predictions.

Bat wings evolved from grasping, manipulating mammalian hands, and this origin influences the biomechanics of flight in bats in comparison to flight in birds and insects. Therefore, an evolutionary perspective is critical to advancing the comparative biology of flight, and helps distinguish those aspects of flight that are shared in all flying animals and those features that are unique to bats. Low weight, particularly in the wings, is important for all flying animals, but selection for reduced wing mass in bats must interact with aspects of neural control in the most morphologically complex of animal wings. In addition, the nature of wing skin as a complex functional material and the capacity to modulate wing mechanical properties during flight by an unusual group of muscles found only in bats proves critical to bat flight performance. Improved understanding of the functional architecture of bat wings not only provides insight into steady-state flight behaviors, but also holds promise for solving problems concerning bats鈥� abilities to recover from perturbations, fly effectively even following wing damage or injury, etc. This approach requires sophisticated bioengineering techniques such as particle image velocimetry, multi-camera high speed videography, and dynamic modeling, but also low-tech methods including polarized light photography, histology, and anatomical description.

Social insects are capable of solving complex physiological problems using collective strategies. I will discuss our work on some of these problems that include the physiology and morphogenesis of termite mounds, and active mechanisms for ventilation, mechanical adaptation and thermoregulation in bee aggregates.

Directed crawling motility of animal cell types ranging from neurons to macrophages requires the coordinated force-generating activity of multiple mechanical elements. Much molecular detail is now known about the constituents of some mechanical submachines such as the polymerizing actin network and the adhesion complexes, but it is not yet clear how these elements all work together to generate coherent, directed motion at the level of the whole cell. In order to understand cellular mechanisms of large-scale coordination, our work focuses on two extremely fast-moving cell types, the fish epidermal basal keratocyte responsible for the rapid closure of wounds in fish skin, and the human neutrophil that hunts down and kills microbial invaders. Despite their very different biological roles and apparent behaviors, these cell share fundamental biophysical mechanisms of self-organization and movement coordination.

Dryland landscapes show a variety of vegetation pattern-formation phenomena, two striking examples of which are banded vegetation on hill slopes and nearly hexagonal patterns of bare-soil gaps in grasslands (鈥渇airy circles鈥�). Vegetation pattern formation is a population-level mechanism to cope with water stress, that is coupled to other response mechanisms operating at lower and higher organization levels, such as phenotypic changes at the organism level and biodiversity changes at the community level. Uncovering the roles that vegetation pattern formation plays in the functioning of dryland ecosystems is a challenging problem of particular significance in the current era of climate change and massive human intervention in natural ecosystems. In this talk I will present a platform of mathematical models for dryland ecosystems and use it to study (i) mechanisms of vegetation pattern formation, (ii) the variety of extended and localized patterns that can appear along the rainfall gradient, (iii) the impact of pattern formation on ecosystem response to droughts, and (iv) forms of high-integrity human intervention that do not impair ecosystem function. Universal aspects that might be applicable to other living systems will be emphasized.

Over the last few decades, new mathematical techniques have played an important role in a variety of biomedical and biophysical imaging modalities. Following a brief survey of the field, we will focus on recent progress in nonlinear optimization that is bringing large-scale problems in acoustic scattering and cryo-electron microscopy within practical reach.

I will present our efforts to map cellular differentiation hierarchies in zebrafish and xenopus embryos, with further examples in mammalian tissues. A few years ago we (and others) developed droplet-microfluidic technology for single cell RNA-Seq (inDrop), opening up the possibility to infer cell states in an unbiased manner. We recently carried out single-cell RNA sequencing of >200,000 cells from time series of the first 24 hours of life, in two vertebrate species: zebrafish (Danio rerio) and the frog (Xenopus tropicalis). We reconstructed the first tens of cell fate choices in the embryo from axis patterning, germ layer formation, and early organogenesis, and compared the two organisms in their differentiation dynamics. We tested how clonally-related cells traverse these fate choices by developing a transposon-based barcoding approach (鈥淭racerSeq鈥�) for reconstructing single-cell lineage histories. Through different examples, I will try to show some of the challenges and opportunities of single cell RNA-Seq approaches: finding and validating new cell types, identifying new growth factor regulators of fate choice, clarifying points of fate commitment of tissues, but also showing where single cell RNA-Seq data can be misleading.

Complex mathematical models of interaction networks are routinely used for prediction in systems biology. However, it is difficult to reconcile network complexities with a formal understanding of their behavior. I will introduce several models of immune recognition by T cells and will show how a simple procedure can be used to reduce them to functional submodules, using statistical mechanics of complex systems combined with a fitness-based approach inspired by in silico evolution. Our procedure works by putting parameters or combination of parameters to some asymptotic limit, while keeping (or slightly improving) the model performance, and requires parameter symmetry breaking for more complex models. An intractable model of immune recognition with close to a hundred individual transition rates is reduced to a simple two-parameter model, and connected to the “adaptive sorting” principle that we previously identified and experimentally validated. Our procedure extracts three different mechanisms for early immune recognition, and automatically discovers similar functional modules in different models of the same process allowing for model classification and comparison.

Swimming cells, including bacteria, sperm, and plankton, play integral roles in processes ranging from human reproduction to ecosystem dynamics to bioremediation in soil. In particular, swimming cells ubiquitously live in dynamic fluid environments that are characterized by complex porous microstructure. In this talk, we will examine the physical mechanisms underlying the transport of motile bacteria in a model microfluidic porous medium, which is used to precisely control both microstructure and flow. We show that such confined flows generate striking heterogeneity in the spatial distribution of motile bacteria due to interactions between cell shape and fluid flow. Unlike passive Brownian particles, the effective diffusion of actively swimming 鈥榬un-and-tumble鈥� bacteria is significantly hindered in flow. In the case of magnetotactic bacteria – which use Earth鈥檚 magnetic field for navigation – we demonstrate that porous media flows can entirely halt their directed motility, locally trapping cells in the porous microstructure. This work illustrates how the physical environment impacts fundamental survival strategies of swimming cells and carries potential implications for biofilm formation and resource competition biomes.

Approximately one-third of global carbon-fixation occurs in an overlooked algal organelle called the pyrenoid. The pyrenoid contains the CO2-fixing enzyme Rubisco, and enhances carbon-fixation by supplying Rubisco with a high concentration of CO2. Since the discovery of the pyrenoid over 130 years ago, the molecular structure and biogenesis of this ecologically fundamental organelle have remained enigmatic. To advance our understanding of the pyrenoid and of photosynthetic organisms more broadly, we have developed new tools for the unicellular model alga Chlamydomonas reinhardtii. These tools include the world鈥檚 first genome-wide collection of mapped mutants in any single-celled photosynthetic organism, as well as methods for high-throughput localization of proteins and identification of protein-protein interactions. By applying these tools, we have increased the number of known pyrenoid components from 6 to over 80, and discovered the existence of three new protein layers in the pyrenoid- a plate-like layer, a mesh layer, and a punctate layer. We discovered that an abundant pyrenoid protein, Essential Pyrenoid Component 1 (EPYC1), works as a molecular glue that binds Rubisco holoenzymes together to form the matrix at the core of the pyrenoid. Furthermore, contrary to longstanding belief that the pyrenoid matrix is a solid structure, we have discovered that the matrix behaves as a liquid droplet, which mixes internally, divides by fission, and dissolves and condenses during the cell cycle. Our efforts have provided fundamental insights into pyrenoid protein composition, structural organization and biogenesis. Working with our collaborators in the Combining Algal and Plant Photosynthesis project, we aim to transfer algal pyrenoid components into higher plants to enhance carbon fixation and yields in crops.

Chromatin structure and dynamics control all aspects of DNA biology yet are poorly understood. In interphase, time between two cell divisions, chromatin fills the cell nucleus in its minimally condensed polymeric state. Chromatin serves as substrate to a number of biological processes, e.g. gene expression and DNA replication, which require it to become locally restructured. These are energy-consuming processes giving rise to non-equilibrium dynamics. Chromatin dynamics has been traditionally studied by imaging of fluorescently labeled nuclear proteins and single DNA-sites, thus focusing only on a small number of tracer particles. Recently, we developed an approach, displacement correlation spectroscopy (DCS) based on time-resolved image correlation analysis, to map chromatin dynamics simultaneously across the whole nucleus in cultured human cells [1]. DCS revealed that chromatin movement was coherent across large regions (4鈥�5渭m) for several seconds. Regions of coherent motion extended beyond the boundaries of single-chromosome territories, suggesting elastic coupling of motion over length scales much larger than those of genes [1]. These large-scale, coupled motions were ATP-dependent and unidirectional for several seconds. Following these observations, we developed a hydrodynamic theory [2] as well as a microscopic model [3] of active chromatin dynamics. In this work we investigate the chromatin interactions with the nuclear envelope and compare the surface dynamics of the chromatin globule with its bulk dynamics [4].

[1] Zidovska A, Weitz DA, Mitchison TJ, PNAS, 110 (39), 15555-15560, 2013

[2] Bruinsma R, Grosberg AY, Rabin Y, Zidovska A, Biophys. J., 106 (9), 1871-1881, 2014

[3] Saintillan D, Shelley MJ, Zidovska A, https://www.biorxiv.org/content/early/2018/05/10/319756, 2018

[4] Chu F, Haley SC, Zidovska A, PNAS, 114 (39), 10338-10343, 2017

The palette of tools for stimulation and regulation of neural activity is continually expanding. One of the new methods being introduced is Magneto-Genetics, where thermo-sensitive and mechano-sensitive ion channels are genetically engineered to be closely coupled to the iron-storage protein ferritin. Such genetic constructs provide a powerful new way of non-invasively activating ion channels in-vivo using external magnetic fields that easily penetrate biological tissue. Initial reports that introduced this new technology have sparked a vigorous debate on the plausibility of physical mechanisms of ion channel activation by means of external magnetic fields. I will argue that the initial criticisms leveled against Magneto-Genetics as being physically implausible were based on the overly simplistic and unnecessarily pessimistic assumptions about the magnetic spin configurations of iron in ferritin protein, and did not comprehensively consider all possible magnetic-field-based mechanisms of ion channel activation in Magneto-Genetics. I will present and propose several new magneto-thermal and magneto-mechanical mechanisms of ion channel activation by iron-loaded ferritin protein, and suggest experiments on a single iron particle-ferritin protein level that may elucidate and clarify some of the mysteries that presently challenge our understanding of the reported biological experiments.

Various challenges are faced when microorganisms or artificial self-propelled particles move autonomously in aqueous media at low Reynolds number. These active agents are intrinsically out of equilibrium and exhibit peculiar dynamical behavior due to the complex interplay of directed swimming motion and stochastic fluctuations. An intriguing feature displayed by microswimmers is the randomization of their swimming motion at large length scales due to reorientation mechanisms such as rotational diffusion or instantaneous tumbling typical of flagellated bacteria. In this talk, I will present recent theoretical advances in the analysis of the spatiotemporal dynamics of different types of active agents in terms of the experimentally measurable intermediate scattering function. Our analytical predictions fully characterize experimental observations of catalytic Janus particles, a paradigmatic class of synthetic active agents, from the smallest length scales where translational Brownian motion dominates, up to the largest ones, which probe the randomization of the swimming direction due to rotational diffusion. Moreover, we elucidate the run-and-tumble motion of E. coli bacteria in terms of a renewal theory. I will also show that our theoretical framework finds application in polymer physics and present our results on the elastic behavior of a semiflexible polymer under compression.

The CRISPR associated Cas9 nuclease has emerged as an exciting new tool for genome-wide pooled forward genetic screens. I will describe our labs efforts to adapt such approaches to study protein quality control pathways and neurodegenerative disease. Specifically, the development of sensitive fluorescence readouts to follow aberrant protein processing, expression, stability and aggregation associated with many of those diseases and combining it with marker-based pooled CRISPR screens for therapeutic target discovery.

Do people who speak different languages think differently? Do languages merely express thoughts, or do they secretly shape the very thoughts we wish to express? Are some thoughts unthinkable without language? Why do we think the way we do? Why does the world appear to us the way it does? Humans communicate with one another using 7,000 or so different languages, and each language differs from the next in innumerable ways. At stake are basic questions all of us have about ourselves, human nature, and reality. I will discuss research conducted around the world and focus on how language shapes the way we think about color, space, time, causality, and agency.

There is an enormous potential for evolution within each of our microbiomes, with billions of new mutations being created each day. In this talk, I will highlight the power of tracking within-person evolution for understanding bacterial transmission, identifying genes critical to long-term survival, and for understanding evolutionary principles. I will present examples from infectious diseases, the gut microbiome, and the skin microbiome. Relevant DOIs: 10.1128/mSystems.00171-17, 10.1101/208009, 10.1038/nm.4205.

The talk will be in two parts. The first part will summarize ongoing work on generating a brain-wide map of brain circuits in the Mouse and the Marmoset at a mesoscopic scale, defined as the transitional scale from individual-variation dominated to a species-typical pattern. The second part will outline some theoretical work on machine learning, and in particular address the puzzling observation that highly overparametrized models seem to still offer good generalization performance. The possibility of a robust crosstalk between an empirical and theoretical research program on intelligent machines will be discussed.

Adaptation is the central evolutionary process and is at the core of some of the greatest challenges facing humanity. Infectious disease, cancer, evolution of resistance to drugs and pesticides are all problems of evolutionary adaptation. Until recently, adaptation was thought to be so idiosyncratic and infrequent that it seemed impossible to obtain enough empirical data about adaptive evolution to investigate it in a systematic way. However, we now know that adaptation is at times sufficiently rapid, recurrent, and involves mutations of such large effect that adaptive dynamics can be quantified on short time scales and with powerful statistical replication. Generally rapid adaptation is a feature of large populations, where the dynamics are not limited by the waiting time for adaptive mutations such that multiple adaptive clones increase in frequency simultaneously, making it difficult to study their individual behaviors. I will describe how we use barcoding and genetic engineering to uncover the dynamics of adaptation in such large populations using experimental evolution in large populations of yeast and in the mouse model of lung cancer.

Metabolism, in addition to being the 鈥渆ngine鈥� of every living cell, plays a major role in the cell-cell and cell-environment relations that shape the dynamics and evolution of microbial communities, e.g. by mediating competition and cross-feeding interactions between different species. Despite the increasing availability of metagenomic sequencing data for numerous microbial ecosystems, fundamental aspects of these communities, such the maintenance of diversity, the unculturability of many isolates, and the conditions necessary for taxonomic or functional stability, are still poorly understood. In our lab, we develop and test mechanistic computational models for the dynamics and evolution of interactions between different organisms based on the knowledge of their entire metabolic networks, with applications in the study of natural and synthetic microbial communities.

Diffusion models, a popular family of decision-making models, are despite their simplicity able to well-describe human and animal decision-making behavior across a surprisingly wide range of different tasks. I will analyze their versatility from a normative perspective, by asking if the behavior they describe is optimal in some sense. This turns out to be the case for a wide range of different tasks that rely on perceptual evidence, and even for decisions that require subjective value judgments, justifying the benefits of acquiring such a decision strategy. Furthermore, it demonstrates that the studied tasks imposed diffusion model-like behavior that might disappear once these tasks become more complex.

Recent progress in systems immunology have ushered a quantitative understanding of T cells鈥� ability to discriminate antigens (e.g. self vs non-self) on a short timescales (&lt 1 hr). Yet modeling how immune responses unfold over long timescales (&gt1 week) remains a challenge with fundamental and clinical applications. Here we will discuss how cell-to-cell communications via cytokine exchange constitute a solution to bridge short and long timescales, local and global responses towards achieving accurate antigen discrimination at the system level. We will present our experimental platform (using CyTOF mass cytometry and a dedicated robotic platform ) to deliver multiplexed time-resolved 鈥渉igh-dimensional鈥� measurements of T cell activation. We will show how we classify complex patterns of T cell activation with machine learning algorithms (top-to-bottom approaches). In parallel, we will discuss how we use biochemical modeling (bottom-up approach) to dissect how signaling cross-talks, synergism and antagonism (between antigen and cytokine response) enable scaling, convergence and divergence of T cell activation.

Soaring birds often rely on ascending thermal plumes (thermals) in the atmosphere as they search for prey or migrate across large distances. How soaring birds find and navigate thermals is unknown. We used reinforcement learning to train gliders to navigate simulated convective turbulent flows. Lessons from simulations allowed us to teach a glider to navigate atmospheric thermals autonomously in the field. Gliders of two-meter wingspan were equipped with a flight controller that precisely controlled the bank angle and pitch, modulating these at intervals with the aim of gaining as much lift as possible. A navigational strategy was determined solely from the gliders鈥� pooled experiences collected over several days in the field. The strategy relies on methods to accurately estimate the local vertical wind accelerations and the roll-wise torques on the glider, which serve as navigational cues. We propose vertical wind accelerations and roll-wise torques as effective mechanosensory cues for soaring birds and provide a navigational strategy applicable to autonomous soaring vehicles.

Highly optimized complex transport networks serve crucial functions in many man-made and natural systems such as power grids and plant or animal vasculature. Often, the relevant optimization functional is nonconvex and characterized by many local extrema. In general, finding the global, or nearly global optimum is difficult. In biological systems, it is believed that such an optimal state is slowly achieved through natural selection. However, general coarse grained models for flow networks with local positive feedback rules for the vessel conductivity typically get trapped in low efficiency, local minima. We show how the growth of the underlying tissue, coupled to the dynamical equations for network development, can drive the system to a dramatically improved optimal state. This general model provides a surprisingly simple explanation for the appearance of highly optimized transport networks in biology such as plant and animal vasculature. In addition, we show how the incorporation of spatially collective fluctuating sources yields a minimal model of realistic reticulation in distribution networks and thus resilience against damage. Distribution networks are shown to exhibit a trade-off between resilience, construction cost, and efficiency, with nature selecting for those phenotypes that lie on the Pareto-efficient front.

The past decades have witnessed a surge in the prevalence of obesity, diabetes and the metabolic syndrome. Many of these disorders are associated with high post-meal blood glucose responses, but common dietary methods for controlling these responses have limited efficacy, mainly due to high interpersonal variability in the response to even the same meal. One of the factors underlying this variability is the gut microbiome: a huge ecosystem of bacteria, archaea, viruses and eukaryotes with vast potential metabolic capacity. In our work we developed new tools for the analysis of the gut microbiome and used these tools, along with blood parameters, dietary habits, anthropometrics and physical activity to accurately predict post-meal blood glucose responses to real-life meals. These predictions were then used to design personalized diets which successfully reduced hyperglycemia. Our results suggest that personalized diets can successfully lower post-meal blood glucose and its grave metabolic consequences.

Spatial localization of chromosomes and their genes is tightly controlled throughout the cell cycle, which ensures proper gene expression during interphase and faithful inheritance of the genome by daughter cells during mitosis. The cellular structures that govern these processes 鈥� the nucleus and the mitotic spindle 鈥� are complex assemblies of biological components, whose mechanical properties are essential to their functions. The two main mechanical components of the cell nucleus are chromatin and the lamina. Corresponding to these components, my coarse-grained polymer model predicts that the nucleus has two mechanical regimes of force response to physical deformations. This picture is supported by micromanipulation experiments in which isolated nuclei are stretched and perturbed through a variety of cell biological and biochemical perturbations. Altogether, this force response observed in the experiments and explained by the model protects the global organization of genome during interphase. During mitosis, after the nucleus disassembles, however, the microtubule spindle governs chromosome positioning. Building on observations from in vitro experiments, I developed a stochastic model for the collective dynamics of microtubules comprising the spindle. The model revealed a striking mechanism for coordinating force-sensitive filament dynamics 鈥� a bistable force-velocity relation. Within this model, I capture known metaphase behaviors such as chromosome oscillations and error correction, and I accurately predict the results of novel experiments in which microtubule dynamics are perturbed via Aurora B kinase. The model thus provides a basic framework for physically understanding regulated metaphase chromosome motions as well as defective motions, which can lead to harmful scenarios, such as aneuploidy. Together, these models show how cellular machinery can robustly control the spatiotemporal positions of chromosomes through structure, geometry, and mechanotransduction.

Bacteria were among the first life forms on Earth, and are found in all of its corners. In order to survive in a variety of environments, bacterial species have developed a variety of locomotive strategies. The most common of these is flagellated swimming, in which bacteria are propelled by the motion of several long filaments that sprout of the cell body. A flagellar filament is rotated by a molecular machine at its base, the aptly-named bacterial flagellar motor (BFM). The motor鈥檚 central role in bacterial motility has made uncovering its operating principles a fundamental challenge in biophysics and microbiology. In this talk, I will present a model for the BFM’s fundamental torque-generation mechanism, and show that model predictions are consistent with experimental observations of motors in various external conditions. I will further discuss recent theoretical and experimental advances in our mechanistic understanding of this nanomachine, and the implications of these advances to several essential biological processes such as bacterial pathogenicity, chemotaxis, and biofilm formation.

Thermal soaring by birds and olfactory searches by insects are biological examples of navigation in the presence of orientation cues that are complex due to the physics of fluids. The two problems also have technological applications, namely for extending the autonomy of flying vehicles/gliders, and for the development of olfactory robots. I shall first review the animal behavior, then present the physics of the orientation cues, and finally discuss the corresponding navigation problems.

In this talk I will discuss our work showing that phase transitions play an important role in organizing the contents of living cells. We focus on a class of membrane-less RNA and protein rich organelles, known as RNP bodies, which help control the flow of genetic information within cells. The nucleolus is one such nuclear RNP body, which is important for cell growth and size homeostasis. We’ve shown that a phase transition model explains many features of nucleolar assembly, and that the internal subcompartments of the nucleolus arise from multi-phase coexistence, which may have important consequences for sequential RNA processing. I will also discuss our new “Optodroplet” approach, which enables spatiotemporal control of phase transitions within living cells, allowing us to begin quantitatively mapping intracellular phase diagrams. This approach has begun to yield rich insights into the link between intracellular liquids, gels, and the onset of pathological protein aggregation.

RNAs are emerging as a powerful substrate for engineering gene expression and cellular behavior since they are now known to control almost all aspects of gene expression. As with all biomolecules, RNA function is intimately related to its structure, since RNA can adopt structures that selectively modulate gene expression. Central questions in biology and bioengineering then are: How do RNAs fold inside cells?; and How can we engineer these folds to control gene expression? In this talk, I will present our work at the interface of these two questions and share results that are beginning to uncover design principles for understanding natural RNAs and engineering RNAs for an array of applications.

I will start by presenting our work on engineering RNA molecular switches that control transcription. The desire to uncover design principles for engineering these RNAs motivates our development of SHAPE-Seq, a technology that couples chemical probing with next-generation sequencing and that helps characterize RNA structures on an 鈥榦mics鈥� scale. I will then describe our exciting recent developments in using SHAPE-Seq to help break open one of the frontiers of RNA structure-function relationships by uncovering at nucleotide resolution how RNAs fold cotranscriptionally. Specifically I will highlight new data on uncovering the ligand-dependent folding pathways of riboswitches, and how we are beginning to use these datasets to computationally reconstruct cotranscriptional folding pathways. This new ability is allowing us to ask deep questions about how RNA molecules make regulatory decisions 鈥榦n the fly鈥� during the dynamic process of transcription. By probing the fundamental processes of RNA folding and function, these studies are expected to greatly aid RNA engineering.

The Dirichlet Process is used to estimate probability distributions that are mixtures of an unknown and unbounded number of components. Amino acid frequencies at homologous positions within related proteins have been fruitfully modeled by Dirichlet mixtures, and we have used the Dirichlet Process to construct such distributions. The resulting mixtures describe multiple alignment data substantially better than do those previously derived. They consist of over 500 components, in contrast to fewer than 40 previously, and provide a novel perspective on protein structure. Individual protein positions should be seen not as falling into one of several categories, but rather as arrayed near probability ridges winding through amino-acid multinomial space.

Quantitative characterization of plasticity, robustness, and evolvability is one of the most important issues in biology. Based on statistical physics and dynamical-systems theory, we present a macroscopic theory of fluctuation and responses in cellular states. By assuming that cells undergo steady growth, protein expression of thousands of genes is shown to change along a one-dimensional manifold in the state space in response to the environmental stress. This leads to a macroscopic law that cellular-state changes satisfy, as is confirmed by adaptation experiments of bacteria under stress. Next, we present proportionality between phenotypic changes by genetic evolution and by environmental adaptation, uncovered both in bacterial experiments and simulations. This relationship is then formulated by the hypothesis that phenotypic changes in adaptation and evolution are dominantly constrained along one-dimensional path. Possible extension of the theory to non-growing cellular states and to multi-level evolution for multicellularity will be briefly discussed.

References
1. Kaneko K., Life: An Introduction to Complex Systems Biology, Springer (2006)
2. K. Sato, Y,Ito, T.Yomo, K. Kaneko, “On the Relation between Fluctuation and Response in Biological Systems” Proc. Nat. Acad. Sci. USA 100 (2003) 14086-14090
3. K. Kaneko, “Evolution of Robustness to Noise and Mutation in Gene Expression Dynamics” PLoS One(2007) 2 e434
4. K. Kaneko, “Phenotypic Plasticity and Robustness: Evolutionary Stability Theory, Gene Expression Dynamics Model, and Laboratory Experiments”, Evolutionary Systems Biology (2012) (Springer, ed. O. Soyer)
5. K. Kaneko, C.Furusawa, T. Yomo, “Macroscopic phenomenology for cells in steady-growth state”, Phys.Rev.X(2015) 011014
7. C. Furusawa, K. Kaneko “Global Relationships in Fluctuation and Response in Adaptive Evolution”, J of Royal Society Interface (2015)

Cancer progression, relapse and resistance are the result of an evolutionary optimization process. Vast intra-tumoral diversity provides the critical substrate for cancer to evolve and adapt to the selective pressures provided by effective therapy. Thus, understanding intra-tumoral diversity and evolutionary dynamics will be a critical step in the development of effective, curative therapies for cancer.

In order to study these questions, we characterized the intra-tumoral genetic heterogeneity of chronic lymphocytic leukemia (CLL) using massively parallel sequencing of large patient cohorts (Landau et al, Cell, 2013, Nature, 2015). These studies have shown that CLLs contain genetically distinct subpopulations that compete and mold the genetic makeup of the malignancy. Furthermore, we have demonstrated that this heterogeneity can help predict the future evolutionary trajectories of CLL, and that higher intra-tumoral heterogeneity in the pre-treatment sample predicts adverse outcome.

Ongoing efforts are dedicated to studying the quantitative evolutionary dynamics that enable the relapse clone to replace the pre-treatment clone (Burger et al, Nature Communications, 2016, and manuscript in review). Using deep sequencing with high temporal resolution we determine the therapy specific clonal fitness with first line chemoimmunotherapy and targeted therapy. These investigations offer a novel framework for the study of the evolutionary dynamics that underlie disease relapse, directly in patients. Moreover, clonal dynamics provide precision prognostication of the timing and clonal composition of relapse. We also use these measurements to tailor personalized therapeutic regimens that would result in longer remissions, laying the foundations for algorithmically driven cancer therapy.

Additionally, in order to comprehensively study cancer evolution, we developed tools to study intra-tumoral epigenetic heterogeneity, as epigenetic somatic changes are heritable and impact the cellular fitness that is selected in the evolutionary process. With these tools, we uncovered a central feature of the cancer epigenome: massive stochastic disorder in methylation patterns. We have further shown that this stochastic disorder impacts transcription, evolution and clinical outcome (Landau et al, Cancer Cell, 2014). Thus, methylation changes in cancer may be similar to the process of genetic diversification, in which stochastic trial and error leads to rare fitness enhancing events. Furthermore, we have performed large-scale single-cell bisulfite sequencing of CLL cells and B cells from healthy donors. We found a close relationship between epimutation and the evolutionary age of the cells. As each generation has a given likelyhood of generating additional stochatic DNA methylation changes, stochastic disorder estimate may reflect the number of generations in the cells evolutionary history. Finally, the phased single cell data allows to reconstruct phylo(epi)genetic relationship between the cells, and infer the stochastic epimutation rate across the genome.

Collectively, these studies demonstrate the tremendous degree of intra-tumoral diversity that fuels cancer evolution, and highlight the need to integrate intra-tumoral heterogeneity in the development of the next generation of cancer therapeutics.

Much of complex biology results from interactions among a large number of individually simpler elements. Behavior of large collection of cells from microbes to stem cells are no different. Nonetheless, the population dynamics of heterogeneous populations is only now beginning to attract attention it deserves because we have only just, within in the last decade or so developed experimental tools for tracking heterogeneous populations. In this talk I will describe how theoretical ideas from statistical mechanics are being used to understand behavior of such heterogeneous populations, focusing on two examples. In first, I will present a coarse-grained model of blood regeneration, which provides a framework to understand large variations (~3 orders of magnitude) among contributions from individual stem cells without active competition. In contrast, the second describes how competition plays a central role in understanding dynamics of reprogramming population of somatic cells.

Protein pattern formation is essential for spatial organization of many intracellular processes like cell division, flagellum positioning, and chemotaxis. More generally, these systems serve as model systems for self-organization, one of the core principles of life.

We present a rigorous theoretical framework able to generalize and unify pattern formation for quantitative mass conserving reaction-diffusion models. Mass redistribution controls local chemical equilibria. Separation of diffusive mass redistribution on the level of conserved species provides a general mathematical procedure to decompose complex reaction-diffusion systems into effectively distinct functional units, and to reveal the general underlying bifurcation scenarios. We apply this framework to Min protein pattern formation and identify the mechanistic roles of both involved protein species. MinD generates polarity through phase separation, whereas MinE takes the role of a control variable regulating the existence of polarized MinD patterns. Hence, polarization and not oscillations is the generic core dynamics of Min proteins in vivo. This establishes an intrinsic mechanistic link between the Min system and a broad class of intracellular pattern forming systems based on bistability and phase separation (wave-pinning). Oscillations are facilitated by MinE redistribution and can be understood mechanistically as relaxation oscillations of the polarization direction.

We demonstrate that the integration of data-driven dynamical systems and machine learning strategies are now capable of extracting governing laws from time-series measurements of physical/biophysical systems. Specifically, we demonstrate that we can use emerging, large-scale time-series data from modern sensors to directly construct, in an adaptive manner, governing equations, even nonlinear dynamics, that best model the system measured using sparsity-promoting techniques. Recent innovations also allow for handling multi-scale physics phenomenon and control protocols in an adaptive and robust way. The overall architecture is equation-free in that the dynamics and control protocols are discovered directly from data acquired from sensors. The theory developed is demonstrated on a number of example problems. Ultimately, the method can be used to construct adaptive controllers which are capable of obtaining and maintaining optimal states while the machine learning and sparse sensing techniques characterize the system itself for rapid state identification and improved optimization.

This lecture will outline a simple way to understand proteins. Per-residue interaction factors will be introduced and used to describe protein structure and to understand heat sensitive mutants, protein-protein interactions (PPI), protein-small molecule interactions (PSMI), and other phenomena. The per-residue interaction factor is a function of amino acid identity, local structure, and multibody contributions; from it, folding/interaction free energy can be calculated. Despite its simplicity, the method compares remarkably well with all-atom models. Determination of these factors is no more complex than the rules to a common board game. Interaction factor heat maps are an especially convenient way to depict the stabilized core of a protein and how the core leverages exterior hot spots. We will illustrate these and other points with a variety of proteins, including general principles that emerge directly from the model. If time permits, we will examine case studies, such as the BCR-Abl kinase domain, commercial drugs that target this kinase, phosphorylation of the activation loop, the conformational switch, and mutational resistance to inhibitors that target the inactive form, since these and many other features of proteins are readily understood and rationalized by casual inspection with these factors.

The three-dimensional structures of proteins often show a modular architecture comprised of discrete structural regions or domains. Cooperative communications between these regions is important to catalysis, regulation and efficient folding; lack of coupling has been implicated in the formation of fibrils and other misfiling pathologies. How different structural regions of a protein communicate and contribute to a protein’s overall energetics and folding, however, is still poorly understood. Here we use a single-molecule optical tweezers approach to indue the selective unfolding of particular regions of T4 lysozyme and monitor the effect of other regions not directly acted on by force. We investigate how the topological organization of a proven (the order of structural elements along the sequence) affects the coupling and folding cooperatively between its domains. To probe the status of the regions not directly subjected to force, we determine the free energy changes during mechanical unfolding using Crooks’ fluctuation theorem. We pull on topological variants (circular per mutants) and find that the topological organization of the polypeptide chain critically determine the folding cooperativity between domains and thus what parts of the folding/unfolding landscape are explored. We speculate that proteins may have evolved to select certain topologies that increase coupling between regions to avoid areas of the landscape that lead to kinetic trapping and misfolding.

The nucleus is mechanically integrated into cells and tissues through LINC complexes, which couple the nuclear envelope to the cytoplasmic cytoskeleton. Nuclear stiffness correlates with tissue stiffness in part through regulation of lamin A expression. However, it is not clear if the lamin A polymer itself imparts nuclear stiffness, or if the ability of lamin A to tether DNA, particularly heterochromatin, to the nuclear envelope (NE) plays a role. Yeast, which lack a nuclear lamina, provide a model system in which to study how tethering of chromatin to the NE influences nuclear stiffness; in this model integral inner nuclear membrane (INM) proteins act as the sole NE tethers. Using a quantitative imaging platform to measure 3D nuclear contours in live cells and an in vitro optical tweezers assay, we found that INM protein tethers promote nuclear stiffness but also restrict chromatin flow to prevent long-term changes in nuclear shape in response to cytoskeletal forces. Interestingly, we find that cells lacking the heterochromatin binding protein HP1 have reduced nuclear stiffness in vivo and in vitro, while driving heterochromatin spreading by deleting the histone H3K9 demethylase makes nuclei more resistant to deformation. This work provides a new framework for thinking about the age-old observation that heterochromatin is constitutively associated with the nuclear envelope in most eukaryotes. In mice, disrupting the ability of cytoskeletal forces to be exerted on chromatin in the nucleus through LINC complex ablation leads to altered differentiation in the skin in vivo and isolated keratinocytes in vitro. These data suggest that physical forces from cell-extracellular matrix (ECM) adhesions are transmitted directly onto the nuclear lamina through LINC complexes to regulate cell fate.

The emergence of spatial patterns in developing multicellular organisms relies on positional cues and cell-cell communication. Drosophila sensory organs have informed a paradigm where these operate in two distinct steps: prepattern factors drive localized proneural activity, then Notch-mediated lateral inhibition singles out neural precursors. Using a combination of experiments and modeling, we show that Notch signaling also organizes the proneural stripes that resolve into rows of sensory bristles on the fly thorax. Patterning, initiated by a gradient of Delta ligand expression, progresses through inhibitory signaling between and within stripes. Thus Notch signaling can support self-organized tissue patterning as a broad prepattern is transduced by cell-cell interactions into a refined arrangement of cellular fates.

The cytoskeleton drives many essential processes in vivo, but for this the system of filaments will arrange itself into different overall spatial organizations, e.g., random, branched networks, parallel bundles, antiparallel arrays, etc. A general objective of our research is to understand what makes these architectures adapted to their tasks. In this talk, I will first focus on 2D disorganized actin networks in which the filaments are oriented randomly in all directions, and are connected both by active molecular motors and passive crosslinkers. Systems with these properties have been reconstituted in vitro, and serve as a model of the cortical actomyosin networks that drive morphogenesis in animal tissues. Although the network components and their properties are known, the requirements for contractility are still poorly understood. I will describe a theory that predicts whether an isotropic network will contract, expand, or conserve its dimensions, depending on the properties of the filaments and the elements that connect them. The theory is simple and encompasses mechanisms of contractions previously proposed.

Julio Belmonte, Francois Nedelec and Maria Leptin
Cell Biology and Biophysics, EMBL Heidelberg, Germany

In this talk I will discuss our work showing that phase transitions play an important role in organizing the contents of living cells. We focus on a class of membrane-less RNA and protein rich organelles, known as RNP bodies, which help control the flow of genetic information within cells. The nucleolus is one such nuclear RNP body, which is important for cell growth and size homeostasis. We’ve shown that a phase transition model explains many features of nucleolar assembly, and that the internal subcompartments of the nucleolus arise from multi-phase coexistence, which may have important consequences for sequential RNA processing. I will also discuss our new “Optodroplet” approach, which enables spatiotemporal control of phase transitions within living cells, allowing us to begin quantitatively mapping intracellular phase diagrams. This approach has begun to yield rich insights into the link between intracellular liquids, gels, and the onset of pathological protein aggregation.

Thermal convection is ubiquitous in nature and can be found in our everyday lives. This subject has been studied by scientists and engineers for many decades, for its rich dynamics and vast applications. In this talk, I will first discuss an experiment as a free-moving floating boundary interacts with a fluid, which is heated from under and cooled from above. The top boundary is mobile and thermally opaque (poor conductor), causing the coupled system to oscillate in space and time. The underlying mechanism is similar to what has been powering the geophysical phenomenon of continental drift, as continents interact with the convective mantle of the earth. In the second experiment, a few seemingly impeding partitions or dividers are inserted into a convective fluid, but the heat-flux that passes through are found to be boosted by several times. Theses results are explained and some new directions that extend the classical picture of thermal convection are also discussed.

Asymmetric cell division provides a means for partitioning not only cell fate determinants but also determinants of cellular age. Increasing evidence suggests that stem cells employ such mechanism for continual maintenance of high proliferative potential. The budding yeast provides an outstanding single-cellular eukaryotic model for study asymmetric aging. Each division cycle of yeast produces a new born bud and an aged mother cell. Two types of aging determinants are segregated asymmetrically: beneficial components that slowly deteriorate but are scarcely replenished, and toxic components that accumulate as a result of acute or chronic cellular stress. Different cellular mechanisms have been discovered to govern the partitioning of these aging determinants. In addition, cellular organelles such as mitochondria play previously unknown roles in cellular quality control and aging.

Cytoskeletal mechanics, in particular the spatial and temporal regulation of force generation and adhesion, regulate a diverse array of physiological processes. While contractility is known to be largely RhoA dependent, the process by which localized biochemical signals are translated into cell-level responses is poorly understood. Here we combine optogenetic control of RhoA signaling, live-cell imaging, and traction force microscopy to investigate the dynamics of actomyosin-based force generation and to measure the local mechanical properties of the cytoskeleton. Local activation of RhoA not only stimulates local recruitment of actin and myosin but also increases traction forces that rapidly propagate across the cell via stress fibers and drive increased actin flow. Surprisingly, this flow reverses direction when local RhoA activation stops. We identify zyxin as a regulator of stress fiber mechanics, as stress fibers are fluid-like and do not exhibit flow reversal in its absence. Using a physical model, we demonstrate that stress fibers behave elastic-like, even at timescales exceeding the turnover of constituent proteins in the cytoskeleton. Such molecular control of actin mechanics likely plays critical roles in regulating morphodynamic events.

Mpemba effect is a counter-intuitive relaxation phenomenon that occurs when, upon coupling to a cold thermal reservoir, a system prepared at a hot temperature relaxes to equilibrium faster than a system prepared at an intermediate temperature. Here we study the possibility of the Mpemba effect for a large class of thermal relaxation processes. We show that an enhanced Mpemba effect is possible for some choices of the initial temperatures, decreasing considerably the cooling times, and introduce a topological index that identifies the existence of such enhancement. We estimate the probability of the Mpemba index for a class of systems with random barriers and energies and show that it is non-vanishing.

Cytokinesis is the physical division of the cell into two daughter cells. Cytokinesis results from the assembly and closure of a ring of actin filaments (F-actin) and the motor protein, myosin-2 at cell cortex (ie the outermost part of the cell). Formation of this ring is controlled by the small GTPase Rho, which itself is activated in a ring-like pattern at the cortex by signals from the mitotic spindle and which, when active, promotes F-actin and myosin-2 assembly. We find that prior to developing the ring-like concentration of Rho activity, the cortex displays propagating, undamped waves of Rho activity and F-actin assembly. Surprisingly, during wave propagation, while Rho elicits F-actin assembly, F-actin subsequently inactivates Rho. Experimental and modeling results show that waves represent excitable dynamics of a reaction-diffusion system with Rho as the activator and F-actin the inhibitor. Ect2, a protein that activates Rho and is required for cytokinesis, stimulates cortical excitability, while spindle microtubules inhibit cortical excitability, thereby confining the waves to the equatorial cortex. Manipulation of MgcRacGAP, a putative Ect2 activator and Rho inactivator, causes “jumping” Rho waves that, rather than steadily propagating, appear and disappear as they progress through the cell cortex. We conclude that the cell cortex is an excitable medium and that the excitability is carried not by ions and ion channels, as in other biological examples of excitability but rather by the cytoskeleton and its regulators. We further conclude that cortical excitability is harnessed by the cell to direct cytokinesis.

Life on Earth constitutes the most sophisticated iterations in the known universe of what physicists classify as soft matter. Research in my group focuses on learning the physical rules of soft matter self-assembly phenomena via the evolutionary processes by which they arose over Earth鈥檚 history. In this view of life as soft matter, evolution, with its own formal rules and algorithms, governs the appearance and diversification of novel forms of soft matter. The field of soft matter was until very recently restricted to analytical consideration of simpler systems like isotropically interacting colloids and cross-linked polymers such as rubber. Our approach allows us to understand soft materials in a nuanced manner that would be inaccessible from more top-down analytical approaches. In this colloquium, I will present the most detailed test case of this perspective to date: the evolutionary appearance of spherical, gradient-index lenses in squids. This complex optical material, first described in theory by Maxwell in 1854, emerges from 5-nm spheroidal proteins via patchy colloidal physics. The lens requires stable, transparent materials throughout the span of packing fractions (from near zero to near one); accordingly, the lens proteins exploit the entire patchy colloidal phase space, and our work is the first demonstration of many of these colloidal organizations in nature. The self-assembling squid proteins exhibit structural nuances that have also been predicted by self-assembly theories, such that the evolved system may provide helpful insight to engineers designing systems at similar lengthscales. Conceptually related projects such as the structure and function of quasi-ordered optics for camouflage of midwater squid eyes will also be discussed.

We discuss some recent experiments and theoretical ideas about chromatin dynamics in live cells and show that understanding these experiments requires the introduction of new concepts such as active noise and scale-dependent viscosity. While the statistical properties of this active, ATP-dependent noise are unknown at present, we argue that anomalous results on dynamics of labelled telomeres in lamin-A-depleted nuclei suggest that this active noise is correlated over very long time scales.

The goal of our research is to elucidate the mechanisms that govern selective filtering by mucus, an important biological hydrogel which coats wet surfaces in the body of all animals. Mucus has critical, but poorly understood, biological functions in protecting tissues from attack by pathogens, and facilitating transport of particulate material. I will present our work on basic mechanisms by which mucus barriers exclude, or allow passage of different molecules and pathogens, and the mechanisms pathogens have evolved to penetrate mucus barriers. We hope to provide the foundation for a theoretical framework that captures general principles governing selectivity in mucus, and likely other biological hydrogels such as the extracellular matrix, and bacterial biofilms. Our work may also be the basis for the reconstitution of synthetic gels that mimic the basic selective properties of biological gel-based barriers.

Nature finds the means to leverage complex geometric and topologic effects in many ways that are we only now beginning to understand. For example, in the case of topology, natural transport webs are frequently dominated by dense sets of nested cycles; the architecture of these networks — the topology and edge weights — determines how efficiently the networks perform their function. We present a new characterization of these physical networks that rests on an abstraction of a physical tiling in the case of a two dimensional network to an effective tiling of an abstract surface in space that the network may be thought to sit in. This new algorithmic approach can be used for automated phenotypic characterization of any weighted network whose structure is dominated by cycles, such as, for example, mammalian vasculature in the organs, the root networks of clonal colonies like quaking aspen, and the force networks in jammed granular matter. On the geometric side of the ledger, it has recently been more and more appreciated that developing biological systems employ complicated 2D stress fields during early onset of morphogenesis from flat or quasi-flat epithelial sheets to a rich zoo of fully three dimensional objects. We discuss a speculative approach based on methods from the physics of exotic shape-shifting materials to reduce the complexity of the interacting “parts” of the stress distribution to model these developmental morphomechanics in a parameter space of drastically reduced dimensionality.

Droplet microfluidics, in which micro-droplets serve as individual reactors, has enabled a range of high-throughput biochemical processes. Although the physics of single drops has been studied extensively, the flow of crowded drops or concentrated emulsions鈥攚here droplet volume fraction exceeds ~80%鈥攊s relatively unexplored in microfluidics. Ability to leverage concentrated emulsions is critical for further increasing the throughput of droplet applications. Prior work on concentrated emulsions focused on their bulk rheological properties. The behavior of individual drops within the emulsion is not well understood, but is important as each droplet carries a different reaction.

This talk examines the collective behavior of drops in a concentrated emulsion by tracking the dynamics and the fate of individual drops within the emulsion. At the fast flow limit, we show that droplet breakup within the emulsion is stochastic. This contrasts the deterministic breakup in classical single-drop studies. We further demonstrate that the breakup probability is described by dimensionless numbers including the capillary number and confinement factor, and the stochasticity originates from the time-varying packing configuration of the drops. To mitigate breakup, we design novel amphiphilic nanoparticles, and show they are more effective than surfactant molecules as droplet stabilizers.

At the slow flow limit, we observe an unexpected order, where the velocity of individual drops in the emulsion exhibits spatiotemporal periodicity. Such periodicity is surprising from both fluid and solid mechanics point of view. We show the phenomenon can be explained by treating the emulsion as a soft crystal undergoing plasticity, in a nanoscale system comprising thousands of atoms as modeled by droplets. Our results represent a new type of collective order not described before, and have practical use in on-chip droplet manipulation. From the solid mechanics perspective, the phenomenon directly contrasts the stochasticity of dislocations in microscopic crystals, and suggests a new approach to control the mechanical forming of nanocrystals.

*Chaos stands for Crowded droplet breakup HydrodynAmics not Ordered but Stochastic.

Active matter is a term that has come to describe diverse systems from flocking animals to the cytoskeleton of a cell. In this talk I will give an overview of the theoretical paradigmthat unifies these diverse systems and discuss some results from minimal models for self propelled particles and suspension of cytoskeletal filaments.

Metastable motions are believed to inhibit the transport of energy through Hamiltonian, or nearly Hamiltonian, systems with many degrees of freedom. We investigate this question in a very simple model (discrete nonlinear Schroedinger equation) in which the breather solutions that are responsible for the metastable states can be computed perturbatively to an arbitrary order. Then, using a modulation hypothesis, we derive estimates for the rate at which the system drifts along a manifold of periodic orbits and verify the optimality of our estimates numerically.

A newly appreciated mechanism of cellular compartmentalization is the formation of liquid-like, membrane-free organelles. These dynamic compartments appear to form via a process of phase separation. Disordered proteins and in many cases RNAs condense into droplets that have specific material properties and these droplets are thought to spatially and temporally control biochemistry and protein translation. Examples of these structures include a variety of RNP granules and the nucleolus. Work in our lab has identified a key role for mRNA sequence in determining both the selectivity of droplet formation and material structures of the condensates. Recent insights into how mRNA can encode specificity and physical properties of these compartments will be presented.

Decades of research have focused on understanding how biochemical signaling and morphogen gradients establish cell patterns during development and tissue morphogenesis. In our research, we find that geometry can drive organization and pattern formation. Mouse embryonic fibroblasts and human vascular smooth muscle cells behave differently on surfaces with various Gaussian curvatures, with pronounced changes in the organization of stress fibers and migration behaviors.

On cylindrical (zero Gaussian curvature) substrates, long, apical stress fibers within these cells align in the direction of minimum curvature, and a second subpopulation of stress fibers, located beneath the nucleus, aligns in the circumferential direction and is bent maximally. We find dramatic reorganization of the actin cytoskeleton upon activation of Rho, which is associated with increased contractility of the fibers[1]. To probe the effects of finite Gaussian curvature, we design substrates that present domains of positive Gaussian curvature connected smoothly to domains of negative Gaussian curvature. This sphere-with-skirt substrate is comprised of a positive-Gaussian curvature spherical cap surrounded by a negative-Gaussian curvature draping skirt that connects smoothly to a planar surface. The spherical cap and skirt both have principal radii similar to cell length scales. Fibroblasts cultured on this substrate exhibit significant reconfiguration of two subpopulations of stress fibers in response to curvature. Cell migration is also strongly connected to the Gaussian curvature. Cells migrating on skirts approach the cap, repolarize to establish a leading edge in the azimuthal direction, and migrate azimuthally around the feature, nearly perpendicular to the apical stress fiber alignment. This mode of migration differs from cell migration on planar surfaces, in which cells typically move in the same direction as the apical stress fiber orientation. Thus, finite Gaussian curvature not only affects the alignment of stress fiber subpopulations, but also directs migration in a manner that deviates from migration on planar surfaces.

We infer that stress fiber alignment is likely a result of a complex balance between energy penalties associated with stress fiber bending, contractility, and the dynamics of F-actin assembly. With a deeper understanding of how cells sense and respond to macroscale curvature, we may be able to use geometric cues to guide cell behaviors such as migration and contraction.

[1] Bade, N. D., Kamien, R. D., Assoian, R. K., & Stebe, K. J. (2017). Curvature and Rho activation differentially control the alignment of cells and stress fibers. Science Advances, 3(9), e1700150.

Computational recognition of distant sequence homology is a key to studying ancient events in molecular evolution. The better our sequence analysis methods are, the deeper in evolutionary time we can see. A major aim in the field is to improve the resolution of homology recognition methods by building increasingly realistic, complex, parameter-rich models. I will describe current and future research in protein, DNA, and RNA homology search algorithms based on probabilistic inference methods, using hidden Markov models (HMMs) and stochastic context-free grammars (SCFGs). We make these methods available in the HMMER and Infernal software from my laboratory, in collaboration with sequence family databases including Pfam and Rfam.

Neurons are different not only because they have different functions but also because they might have different genealogies. However, the enormous diversity of neurons both within the same nervous system and across species presents tremendous challenges for their unbiased classification. Here, I will discuss novel approaches and algorithms toward establishing the natural classification of neurons. Our research strategy is based upon (1) high-throughput single-cell RNA-seq and single-cell epigenomic analyses of entire neuronal circuits as they learn and remember, and (2) implementation of stochastic approaches from phylogenomics. Our results suggest that different classes of neurons and synapses (as well as complex brains) might have evolved more than once (convergent evolution) and allow us to start reconstruction the genealogy of neurons, trace ancestral cell lineages, and establish the natural classification of neurons within neural circuits across the majority of animal phyla: from ctenophores and cnidarians to bilaterians including cephalopods. The field of Neurosystematics is emerging. This might be an analog of the periodic table for neurons, with the predictive power to delineate novel neuronal phenotypes and fundamental constraints on the origins and parallel evolution of neural systems. In summary, there is more than one way to develop neuronal complexity, and animals frequently use different molecular toolkits to achieve similar functional outcomes.

The properties of amorphous solids are essentially and qualitatively different from those of simple crystals. Unless a crystal’s unit cell is very complicated, all particles or inter-particle bonds contribute nearly equally to any global quantity, so that each bond plays a similar role in determining the physical properties of the solid. For example, removing a single bond in a perfectly ordered array or network decreases the overall elastic strength of the system, but in such a way that the resistance to shear and the resistance to compression drop in tandem so that their ratio is nearly unaffected. Disordered materials are not similarly constrained. We introduce a principle unique to disordered solids wherein the contribution of any bond to one global perturbation is uncorrelated with its contribution to another. Coupled with sufficient variability in the contributions of different bonds, this “independent bond-level response” paves the way for the design of real materials with unusual and exquisitely tuned properties. For example, we can tune a disordered network鈥檚 Poisson ratio anywhere between the auxetic and incompressible limits. We can also produce a targeted response at a local scale; by perturbing the positions of pair of particles at one point we can tune in a desired response a large distance away. This response is reminiscent of allosteric regulation in proteins where a reaction at one site activates another site of the protein molecule. Experimentally, we have successfully demonstrated such mechanical networks in 2D (by laser cutting) or in 3D (3D printing).

Engineered sensors allow us to accurately measure external stimuli, such as temperature, pressure, light, and sound. However, how biological organisms detect some of these same stimuli remains poorly understood. One of the most fundamental of these is the detection of temperature: avoiding cell damage while seeking optimal temperatures for cell physiology is key to the survival of both complex and simple organisms. While temperature affects many rates within the cell, certain components have evolved specifically to function as temperature sensors. While we know these temperature sensors exist in various forms, such as the changing of membrane fluidity, folding of RNA thermometers, or opening of ion channel pores, our understanding of how these distinct mechanisms are invoked by identical changes in temperature is incomplete. Did evolution harness the same statistical mechanical principles for all of these? For some models of temperature sensing, parallels can be drawn to problems in protein folding. This talk will walk through recent developments relevant to understanding temperature sensing in biology. First will be an overview of the exciting advances of the last decade in understanding the TRP channel family: a temperature sensing family of ion channel proteins (present only in eukaryotes), in which hot and cold sensitive channels are also sensitive to the small molecules capsaicin (of chili peppers) and menthol (of mint), respectively. Then we will examine the variety of tools initially developed to perform and analyze long timescale molecular simulations for protein folding applied to model protein systems chosen to refine the statistical mechanical understanding of small molecule binding to proteins. The talk will conclude with a look forward to how we can combine these tools to define model temperature sensing systems and develop a general framework for understanding temperature sensing mechanisms across biological organisms.

For the most part microbial life does not swim through simplebulk Newtonian fluids but rather moves through complex environments. Surfaces, interfaces and confinement are ubiquitous in microbial environments . Motile microbes can be either aided or deterred from reaching surfaces by external flows, and the flowing medium may exhibit a dual fluidic and elastic nature. In this seminar, we will discuss the dynamics and hydrodynamics of self-propelled microorganisms in these biologically relevant environments. In particular, swimmer trajectories within flowing films will be considered in order to determine how swimming strategy results in differing swimmer distributions. It will be shown that the rheology of flowing biological medium plays an important role in setting the ability of microbes to swim upstream in microchannels. We will then turn our attention to suspensions of many swimmers confined within an array of responsive obstacles to explore the ability of collective motion to restructure micro-environments. We model the dense suspensions of motile cells as a complex fluid: an active nematic liquid crystal. Our simulations show that a lattice of rotors immersed a bacterial suspension in self-organizes into a spin-state wherein neighboring disc-like obstacles continuously rotate in permanent alternating directions due to combined hydrodynamic and elastic effects. This antiferromagnetic spin-state is permanent and only exists for sufficiently small inter-disc spacing. The existence of such active matter-mediated forces between passive bodies suggests the ability of living suspensions to facilitate collective motion by altering their surroundings.

Metabolism is the set of enzymatic reactions that cells use to generate energy and biomass. Interestingly, recent studies suggest that many metabolic enzymes assemble into large clusters, often in response to environmental conditions. Theoretically, we find that large-scale enzyme clusters, with no internal spatial ordering of enzymes, offer many of the same advantages as direct substrate channeling: accelerating intermediate processing, protecting intermediates from degradation/cross-reactions, and protecting the cell from toxic intermediates. The model predicts the separation and size of coclusters that maximize metabolic efficiency. For direct validation, we study a metabolic branch point in Escherichia coli and experimentally confirm the model predictions. Our studies establish a quantitative framework to understand coclustering-mediated metabolic channeling and its application to both efficiency improvement and metabolic regulation.

Protein-protein interactions are critical to the operation and functions of all cells. The specificity of these interactions is often dictated at the level of molecular recognition, meaning proteins have an intrinsic ability to discriminate cognate from non-cognate partners. Understanding precisely how this discrimination is accomplished remains a major problem, particularly for paralogous protein families in which the individual members share high sequence and structural similarity. Our work tackles this problem primarily in the context of two-component signal transduction systems, the predominant form of signaling in bacteria, and more recently with toxin-antitoxin systems, also found throughout the bacterial kingdom. I will describe our work using analyses of amino acid coevolution to pinpoint the molecular basis of specificity in these proteins. This work has enabled the rational rewiring of protein-protein interactions and signal transduction pathways. Additionally, these studies have driven efforts to systematically map sequence spaces and probe the selective pressures and constraints that govern the evolution of protein-protein interactions.

Like the human retina, the Drosophila retina contains randomly distributed color photoreceptor cells that are defined by the expression of different color sensitive Rhodopsins. In Drosophila, two types of individual unit eyes (ommatidia) are specified by stochastic expression of the transcription factor Spineless. This decision is controlled by a two-step process: First, each allele of spineless randomly makes a cell-intrinsic ON/OFF expression decision governed by global activation coupled with stochastic repression. When the expression decisions disagree (one allele ON and one allele OFF), interchromosal communication coordinates expression state between the two alleles. This effect does not depend on chromosomal pairing or endogenous spineless chromosomal position but instead requires specific DNA elements to mediate regulatory interactions. This mechanism couples stochastic repression with interallelic coordination and contrasts starkly with the noisy activation mechanisms seen in bacteria, and the mono-allelic, stochastic activation mechanisms observed in the mouse olfactory and human color vision systems.

Many vertebrate and invertebrate eyes also have retinal mosaics that contain different stochastically specified types of photoreceptors. At least one group, the butterflies, make a three-way stochastic choice between three ommatidial types. However, it remains unclear how much of the regulatory network that specifies photoreceptor subtypes is retained or has evolved in other insects, and whether they use stochastic Spineless expression to diversify their sometimes more complex retinal mosaics. I will present evidence that a conserved regulatory code defines and expands photoreceptor subtypes between flies (Drosophila) and butterflies (Papilio xuthus and Vanessa cardui). We used CRISPR/Cas9 to knock out Spineless in butterflies and provide functional evidence that there is deep evolutionary conservation of stochastic patterning mechanisms. Furthermore, butterflies have two R7 photoreceptors that allow for the specification of three types of ommatidia instead of two. This in turn allowed for the evolution and deployment of additional opsins, tetrachromacy, and improved color vision, important features of butterfly life history and ecology. Our extensive knowledge of patterning in the Drosophila visual system applies to other groups, and adaptation for specific visual requirements can occur through modification of this network.

Patterning of the neurons that process visual information in the optic lobes is, in contrast to the retina, highly deterministic. The medulla, where motion and color information are processed, contains 40,000 neurons of more than 70 cell types. These neurons are born from neural stem cells (Neuroblasts) that sequentially express five transcription factors. The neurons emerging from neuroblasts at each stage maintain expression of the corresponding gene and become different cell types. We will describe the mechanisms controlling the transition from one neuroblast stage to the next. The neuroepithelium that generates the medulla neuroblasts is also highly patterned, with eight regions defined by the expression of spatially restricted transcription factors. Each region contributes to generating two types of neurons: 鈥楿ni-columnar neurons鈥� that have a 1:1 stoichiometry with the photoreceptors that innervate the medulla and are generated throughout the neuroepithelium. The less numerous 鈥榥on-columnar鈥� neurons, which contact multiple photoreceptors, are generated from specific sub-regions of the neuroepithelium and migrate to take on their retinotopic position in the medulla. This combination of temporal and spatial patterning allows for the generation of the 70 medulla cell types.

Why do animals move the way they do? Bacteria, insects, birds, and fish share with us the necessity to move so as to live. Although each organism follows its own evolutionary course, it also obeys a set of common laws. At the very least, the movement of animals, like that of planets, is governed by Newton鈥檚 law: All things fall. On Earth, most things fall in air or water, and their motions are thus subject to the laws of hydrodynamics. Through trial and error, animals have found ways to interact with fluid so they can float, drift, swim, sail, glide, soar, and fly. This elementary struggle to escape the fate of falling shapes the development of motors, sensors, and mind. Perhaps we can deduce parts of their neural computations by understanding what animals must do so as not to fall.

In this talk I will discuss recent developments along this line of inquiry in the case of insect flight. Asking how often a fly must sense its orientation in order to balance in air has shed new light on the role of motor neurons and steering muscles responsible for flight stability.

The self-assembly of a protein shell around a cargo is a common mechanism of encapsulation in biology, and is inspiring development of drug delivery vehicles that form by self-assembly. However, the physics underlying such multicomponent assembly processes is incompletely understood. In this talk I will describe how minimal computational models can elucidate two biological examples in which icosahedral protein shells assemble around cargos. In each case we find that the material properties of the cargo play a key role in directing its encapsulation.

The first example concerns viruses with single-stranded RNA (ssRNA) genomes. For many ssRNA viruses, formation of an infectious virus requires the spontaneous assembly of an icosahedral protein shell (called a capsid) around the genome. I will describe simulations that investigate how this co-assembly process depends on the physical properties of RNA: its length, electrostatic charge, and 3D structure. When applied to specific virus capsids, the calculated optimal RNA lengths closely correspond to the natural viral genome lengths. This suggests that evolution of viral RNA is driven not only by the fitness of the proteins that it encodes for, but also by how its material properties favor encapsulation. We then show that assembly can proceed through two qualitatively different classes of pathways, which can be tuned by solution conditions or changing the capsid protein properties.

The second example concerns carboxysomes, which are large, roughly icosahedral protein shells that facilitate carbon fixation in cyanobacteria. Carboxysomes assemble around a cargo which is topologically different from ssRNA, a noncovalently linked, amorphous complex of the enzyme RuBisCO. Motivated by this problem, we study assembly of icosahedral shells around a fluid cargo. We find different assembly pathways and different critical control parameters as compared to assembly around RNA, and that the predominant assembly pathway depends strongly on the cargo fluidity. We discuss relationships between simulated assembly pathways and recent experiments observing assembly of individual carboxysomes in bacteria.

The beating patterns of sperm flagella and the breast-stroke like swimming of ciliates are driven by the molecular motor dynein. This motor generates sliding forces between adjacent microtubule doublets within the axoneme, the motile cytoskeletal structure. To create regular, oscillatory beating patterns, the activities of the dyneins must be coordinated both spatially and temporally. It is thought that coordination is mediated by stresses or strains that build up within the moving axoneme, but it is not known which components of stress or strain are involved, nor how they feed back on the dyneins. To answer this question, we measured the beating patterns of isolated, reactivate axonemes of the unicellular alga Chlamydomonas. We compared the measurements in wildtype and mutant cells with models derived from different feedback mechanisms. We found that regulation by changes in axonemal curvature was the only mechanism that accords with the measurements.

Cells operate in noisy molecular environments via complex regulatory networks. It is possible to understand how molecular counts are related to noise in specific networks, but it is not generally clear how noise relates to network complexity, because different levels of complexity also imply different overall number of molecules. For a fixed function, does increased network complexity reduce noise, beyond the mere increase of overall molecular counts? If so, complexity could provide an advantage counteracting the costs involved in maintaining larger networks.

For that purpose, we investigate how noise affects multistable systems, where a small amount of noise could lead to very different outcomes; thus we turn to biochemical switches like the G2/M cell cycle transition switch. Our method for comparing networks of different structure and complexity is to place them in conditions where they produce exactly the same deterministic function. We are then in a good position to compare their noise characteristics relatively to their identical deterministic traces.

We show that more complex networks are better at coping with both intrinsic and extrinsic noise. Intrinsic noise tends to decrease with complexity, and extrinsic noise tends to have less impact. Our findings suggest a new role for increased complexity in biological networks, at parity of function.

HIV continues to wreak havoc around the world, especially in poor countries. A vaccine is urgently needed to overcome this major global health challenge. I will describe key challenges that must be confronted to achieve this goal. I will then focus on some work that aims to address a part of these challenges by bringing together theory and computation (rooted in statistical physics), consideration of structures of multi-protein assemblies, basic immunology, and human clinical data. The results of these studies suggest the design of immunogens and immunization strategies for vaccines that might elicit immune responses which might be able to hit HIV where it hurts upon natural infection.

Inverse statistical approaches, modeling pairwise correlations between amino acids in the sequences of homologous proteins across many different organisms, can successfully extract protein structure (contact) information. Here, we benchmark those statistical approaches on exactly solvable models of proteins, folding on a 3D lattice, to assess the reasons underlying their success and their limitations. We show that the inferred parameters (effective pairwise interactions) of the statistical models have clear and quantitative interpretations in terms of positive (favoring the native fold) and negative (disfavoring competing folds) protein sequence design. New sequences randomly drawn from the statistical models are likely to fold into the native structures when effective pairwise interactions are accurately inferred, a performance which cannot be achieved with independent-site models.

Impressive progress in membrane biology has revealed important relationships in G-protein coupled receptor (GPCR) structure and function. Despite this, understanding of the largest family of GPCRs, the mammalian odorant receptors (ORs), lags behind. To date, there is no crystal structure of any ORs and many remain orphans without known ligands. These challenges must be overcome in order to understand the molecular mechanisms of olfaction.

We have developed a new strategy that enables comprehensive screening of ORs in freely behaving animals by odor stimulation. We combine phosphorylated ribosomal protein S6 immunoprecipitation with next generation sequencing to profile OR expression in active olfactory sensory neurons. This method is capable of not only identifying a repertoire of odorant-OR pairs, but also reveals the most robust and sensitive ORs. This deorphanization is key to understanding structure-function relationships of odorant-OR interactions.

Poor cell surface expression of ORs in heterologous cells represents a major challenge in functional studies of ORs. We previously identified the RTP family of accessory proteins that facilitate OR cell surface expression, and used them for functional analysis in heterologous cells. A transcriptomic and in situ hybridization analysis of olfactory mucosa of RTP1 and RTP2 knockout mice has revealed that the majority of ORs are downregulated, whereas a small subset of ORs are upregulated. This subset of ORs demonstrate expression at the cell surface in heterologous systems, suggesting an unexpected connection between OR protein trafficking and gene choice.

Understanding the mechanisms of learning and memory is one of the major challenges in neuroscience. The dominant theory holds that information about sensory inputs is stored in cortical circuits thanks to synaptic plasticity. In spite of decades of research, the exact rules governing how synapses change as a function of the activity of pre and post-synaptic neurons remain the subject of debate. In this talk, I will present two novel approaches for investigating the mechanisms of learning and memory. The first consists in inferring a learning rule from in vivo data, using experiments comparing the statistics of responses of neurons to large sets of novel and familiar stimuli. The second consists in exploring the consequences of an information optimization principle on the statistics of synaptic connectivity. I will show how methods from statistical physics can be used to characterize the statistics of connectivity in networks that optimize information storage, and compare the theoretical results with available data.

What do the laws of thermodynamics look like, when applied to microscopic systems such as optically trapped colloidal particles, single molecules manipulated with laser tweezers, and biomolecular machines? In recent years it has become apparent that the fluctuations of small systems far from thermal equilibrium satisfy strong and unexpected laws, which allow us to rewrite familiar inequalities of macroscopic thermodynamics as equalities. These results in turn have spurred a renewed interest in the feedback control of small systems and the closely related Maxwell鈥檚 demon paradox. I will describe some of this progress, and will argue that it has refined our understanding of irreversibility, the second law, and the thermodynamic arrow of time.

I will report on work with Tsvi Tlusty on how one can envisage a concrete map between the genetic information contained in the DNA sequence and mechanical function of the protein.

The simplicity of our model allows for an extensive survey of the “protein universe”, spanning a huge number of generations and samples, far beyond what can be done in real living matter.

This helps in understanding how mechanical constraints on the function of the protein force the system into a very small subset of possible states. Evolution can be described in terms of a low-dimensional random walk on this subset of an almost infinite-dimensional space. This is also the origin of tight correspondence between the spectrum of functional DNA sequences and the mechanical modes of the protein.

Fish keratocyte cells served as the model system to understand biophysics of cell motility for decades. Recently, we combined experiment and modeling to understand the mechanism of polarization of these cells. We found that two essential feedbacks – positive one between myosin density and actin flow, and negative one between stick-slip adhesions and actin flow – underlie the motility initiation. Interestingly, keratocytes polarize in electric fields much faster but not stably, through different mechanism. I will also describe preliminary results on collective keratocyte migration in electric fields.

Coating colloidal particles with DNA is emerging as a new way to direct self assembly. In principle, it allows one to program the assembly of different materials in almost any way imaginable. Here we describe recent progress, which includes the introduction of valence and novel superlattices to create new colloidal structures for photonic applications.

I will present our advances in combining computational and experimental techniques to develop novel inhibitors. We have developed an integrated pipeline that first computationally designs large libraries of potential inhibitors and can then screen these for either cellular phenotype or high affinity binding. I will showcase this pipeline on two example applications, first for developing inhibitors to protein-protein interactions and second for developing novel high-affinity biologics.

Equilibrium fluctuation-induced forces are abundant in nature, ranging from quantum electrodynamic (QED) Casimir and van der Waals forces, to their thermal analogs in fluctuating soft matter. Manifestations of QED fluctuations out of thermal equilibrium are also well-known, as in the Stefan-Boltzmann laws of radiation pressure and heat transfer. These laws, however, acquire non-trivial twists in the near-field regime of sub-micron separations, and in the proximity of moving surfaces. I will discuss dissipation in moving steady states, and the non-Gaussian fluctuations of a particle in a quantum bath.
Non-equilibrium fluctuation forces for particulate matter also hold surprises which I present in the contexts of diffusive transport, and active matter: For the simple case of a current of diffusive particles between parallel slabs, we find a force that falls off with slab separation d as kT/d (at temperature T, and in all spatial dimensions), but that can be attractive or repulsive. There is also a universal transient force when a system of particles undergoes temperature quench or sudden agitation. For a wide wide class of active systems, we find that the pressure exerted on a container depends on details of interactions with the confining walls, as well as wall curvature and asymmetry.

Living organisms need to obtain and process information that are crucial for their survival. Information processing in living systems, ranging from signal transduction in a single cell to image processing in the human brain, are performed by biological circuits (networks of interacting bio-molecules or neurons), which are inherently noisy. However, certain accuracy is required to carry out proper biological functions. How do biological networks process information with noisy components? What is the free energy cost of accurate biological computing? Are there fundamental physics principles underlying the performance of these biological circuits? In this talk, we will describe our recent work in trying to address these general questions in the context of two basic cellular computing tasks: sensory adaptation for memory encoding; biochemical oscillation for accurate timekeeping.

Particles may phase separate because they are of different sizes, of different shapes, or interact differently. But there is also the possibility that they separate based on the different level of activity. I will present a simple model which illustrates this idea. I will also discuss how situation changes if particles are connected to form a polymer chain, and speculate about possible implications of these results.

We’ll review some open source software our lab has developed to facilitate a Phase I clinical trial of a personalized therapeutic vaccine targeting tumor neoepitopes. We’ll also present some results from our analysis of clinical trials of checkpoint blockade in 3 different cancer types and some open source software we’ve created to facilitate similar analyses.

Active materials can self-organize in many more ways than their equilibrium counterparts. For example, self-propelled particles whose velocity decreases with their density can display motility-induced phase separation (MIPS), a phenomenon building on a positive feedback loop in which patterns emerge in locations where the particles slow down. Here, we investigate the effects of intrinsic fluctuations in the system’s dynamics on MIPS, using a field theoretic description building on results by Cates and collaborators. We show that these fluctuations can lead to transitions between metastable patterns. The pathway and rate of these transitions is analyzed within the realm of large deviation theory, and they are shown to proceed in a very different way than one would predict from arguments based on detailed-balance and microscopic reversibility. Specifically, we show that these transitions involve fluctuations in diffiusivity of the bacteria followed by fluctuations in their population, in a specific sequence. The method of analysis proposed here, including its numerical component, can be used to study noise-induced non-equilibrium transitions in a variety of other non-equilibrium set-ups, and leads to predictions that are verifiable experimentally.

Abstract: The characterization of virus populations by deep sequencing is transforming our understanding of viral evolutionary dynamics. This enables us to address questions about the extent of within-host virus diversity and what proportion of this diversity is transmitted between infected hosts. Using the same tools we can also query the host environment in which the virus evolves鈥攕uch as host microbial ecology and local response to infection鈥攖o determine its effect on virus evolution. I will discuss how we quantify virus diversity and characterize virus variants that achieve sustainable transmission, and illustrate how immune status, the respiratory microbiome, and mixed infections can shape influenza virus transmission.

We normally think of evolution occurring in a population of organisms, in response to their external environment. Rapid evolution of cellular populations also occurs within our bodies, as the adaptive immune system works to eliminate infection. Some pathogens, such as HIV, are able to persist in a host for extended periods of time, during which they evolve to evade the immune response. In this talk I will introduce an analytical framework for the rapid coevolution of B-cell and viral populations. I will quantify the amount of out-of-equilibrium adaptation in each of the two populations by analysis of their co-evolutionary history. I will discuss the consequences of competition between lineages of antibodies, and characterize the fate of a given lineage dependent on the state of the antibody and viral populations. In particular, I will discuss the conditions for emergence of highly potent broadly neutralizing antibodies, which are now recognized as critical for designing an effective vaccine against HIV.

Biology entered a new era, with computational biology producing biological data that are impossible nowadays to obtain with wet experiments. Tackling biological questions with advanced engineering, new computer algorithms and novel computational approaches is a challenge that will lead to revolutionize biology and medicine through deeper, ubiquitous use of DNA information.

A fundamental question is the extraction of evolutionary information from DNA sequences. Here, we consider protein sequences and structures, and describe how important biological information on protein-protein binding sites and on mechanical and allosteric properties of proteins can be extracted. Among different examples, we shall present a computational approach to protein-protein interactions that we developed within a project on protein network reconstruction on neuromuscular diseases. The project demands a high computational power to test billions of interactions, it ran on the machines of the World Community Grid for more than 3 years, and provided a huge amount of information on the interaction of human proteins. High Performance Computing helped to obtain an unprecedented amount of information on protein-protein interactions between real partners but also, and most importantly,
between non-partners.

Biological phenomena unfold in a broad range of scales ranging from molecules to cells to populations and ecosystems. Variation of molecular properties of biomolecules profoundly impact the ability of cells to survive and propagate (fitness). Finally, the fate of a mutation is decided by Darwinian selection on the level of the population, where three outcomes are possible: fixation in the population, elimination by purifying selection or separation in the population in a subdominant clone (polymorphism). In this lecture I will outline my lab鈥檚 and others efforts in an emerging new field which merges molecular mechanism with evolution.

I will discuss new models of evolutionary dynamics on biophysical fitness landscapes. Traditional population genetics models are agnostic to the physical-chemical nature of mutational effects. Rather they operate with an a鈥檖riori assumed distributions of fitness effects (DFE) of mutations from which evolutionary dynamics are derived. In departure with this tradition the novel multiscale models integrate the molecular effects of mutations on physical properties of proteins into physically intuitive yet detailed genotype-phenotype relationship (GPR) assumptions. I will present a range of models from simple analytical diffusion-based model on biophysical fitness landscapes to more sophisticated computational models of populations of model cells where genetic changes are mapped into molecular effects using biophysical modeling of proteins and ensuing fitness changes determine the fate of mutations in realistic population dynamics. Examples of insights derived from biophysics-based multiscale models include the fundamental limit on mutation rates in living organisms, physics of thermal adaptation, co-evolution of protein interactions and abundances in cytoplasm and related results, some of which I will briefly present and discuss.

Next, I will present major results from novel 鈥渂ottom-up鈥� experimental approach to study evolutionary dynamics on biophysical fitness landscapes. The approach spans all scales of biological organization involving concurrent use of genome editing, biophysical characterization of molecular effects of mutations, high throughput proteomic analysis at the systems level and phenotypic analysis. It validates and further develops the concept of biophysical fitness landscapes by showing that certain combinations of molecular traits can serve as a universal predictor of fitness effects of mutations. Thus linking fitness effects to intermediate phenotype 鈥� molecular and cellular effects of mutations 鈥� provides a comprehensive low-dimensional mapping of genotype to phenotype 鈥� a biophysical fitness landscape – on which evolutionary dynamics unfolds.

The essence of Structural DNA Nanotechnology is the combination of branched DNA molecules combined with interactions that can be prescribed by Watson-Crick base pairing. The key goals of the area include the production of objects, lattices and nanomechanical devices made from DNA, as well as controlling the positions of other materials. By the middle 1990s, geometrical control was achieved, leading well-defined objects, often objects acting as tiles for 2D lattices.

Nanorobotics is a key area of application. We have made robust 2-state and 3-state sequence-dependent devices and bipedal walkers. We have constructed a molecular assembly line using a DNA origami layer and three 2-state devices, so that there are eight different states represented by their arrangements. All eight products can be built from this system.

Organization of other materials by DNA is another key goal. We have placed differently-sized gold nanoparticles in a checkerboard array in 2D, and in specific positions in 3D. We have also placed carbon nanotubes on DNA origami in specific positions.

There is an empirical rule stating that the best arrays in multidimensional DNA systems result when helix axes span each dimension. We have self-assembled a 2D crystalline origami array by applying this rule. We used the same rule to self-assemble a 3D crystalline array. We initially reported its crystal structure to 4 脜 resolution, but rational design of intermolecular contacts has enabled us to improve the crystal resolution to better than 3 脜. We can use crystals with two molecules in the crystallographic repeat to control the color of the crystals. We can change the color of crystals by doing strand displacement of duplex DNA; we can also color the crystals using triplex formation. When tailed in DNA, we can add semiconductors to the crystals, and follow their transitions by crystal color. The use of the crystals to host guests promises an approach to the organization of macromolecules in 3D. Diffraction of the crystals offers a means to ascertain the successful construction of their targets and the characterization of their guests.

This Research was supported by MURI W911NF-11-1-0024 from ARO, N000141110729 from ONR, grants EFRI-1332411 and CCF-1526650 from the NSF, DE-SC0007991 from DOE for DNA synthesis and partial salary support, and grant GBMF3849 from the Gordon and Betty Moore Foundation.

Cell size is an important physiological trait that is regulated by coupling growth with cell division. Although many of the key regulatory proteins effecting size control have been identified, the underlying molecular mechanisms are unclear. I will discuss our recent progress in addressing this fundamental question in proliferating budding yeast and in developing frog embryos. The common feature of yeast and frog mechanisms is that the synthesis of activating and inhibiting molecules scales differently with cell size. Through this differential size scaling, the ratio of activator and inhibitor becomes a monotonic function of cell size so that the cellular transition becomes size-dependent. Importantly, this mechanism can function in the context of any regulatory network and any cell geometry. Thus, the differential size scaling of macromolecular synthesis provides an elegant mechanism to couple cellular transitions with cell size.

Models of systems biology, climate change, ecosystems, and macroeconomics have parameters that are hard or impossible to measure directly. If we fit these unknown parameters, fiddling with them until they agree with past experiments, how much can we trust their predictions? We have found that predictions can be made despite huge uncertainties in the parameters — many parameter combinations are mostly unimportant to the collective behavior. We will use ideas and methods from differential geometry to explain what sloppiness is and why it happens so often. Finally, we shall show that field-theory models in physics are also sloppy 鈥� that sloppiness makes science possible.

Pressure is the mechanical force per unit area that a confined system exerts on its container. In thermal equilibrium, the pressure depends only on bulk properties (density, temperature, etc.) through an equation of state. The talk will show that in active systems containing self-propelled particles, the pressure instead can depend on the precise interactions between the system’s contents and its confining walls. This implies that generic active systems have no equation of state. Other anomalous attributes of pressure will also be discussed.

Finally, it will be shown that in certain fine tuned cases an equation of state can be recovered. The physics behind the equation of state in a specific example will be discussed.

Cell migration is important in many biological processes, including embryonic development, cancer metastasis, and wound healing. In these tissues, a cell’s motion is often strongly constrained by its neighbors, leading to glassy dynamics.

While self-propelled particle models exhibit a density-driven glass transition, this does not explain liquid-to-solid transitions in confluent tissues, where there are no gaps between cells and therefore the density is constant. Here we demonstrate the existence of a new type of rigidity transition that occurs in the well-studied vertex model for confluent tissue monolayers at constant density. We find the onset of rigidity is governed by a model parameter that encodes single-cell properties such as cell-cell adhesion and cortical tension, providing an explanation for a liquid-to-solid transitions in confluent tissues and making testable predictions about how these transitions differ from those in particulate matter.

The study of chromosomes has a rich history; however, until recently, the internal structure of chromosomes remained largely unexplored. Based on studies using light microscopy, electron microscopy, tomography and mechanical measurements, various models of chromosomes have been proposed. A recent molecular technology Hi-C, introduced in 2009, have enabled studies of chromosomal folding on a genome-wide scale at high resolution, and was since applied to a range of organisms and cell types.

When applied to interphase human cells, Hi-C reveals several levels of compartmentalization of chromosomes. At megabase level, the genome is partitioned into transcribed gene-rich regions, enriched in active chromatin marks (鈥渁ctive鈥� regions), and 鈥渋nactive鈥� gene-poor regions. At sub-megabase level, chromosomes can be viewed as a sequence of Topologically Associated Domains (TADs), which are 100-800kb regions enriched in chromosomal contacts. Finally, specific looping interactions occur within TADs.

We applied Hi-C to human cells in metaphase, a cell state dedicated to division. We found that mitotic chromosomes assume an entirely different folded state, in which all three levels of interphase chromosome compartmentalization are lost. Moreover, the distribution of chromosomal contact changes: more contacts occur at distances 10Mb, and less at larger distances.

Previously, two fundamental classes of models were proposed for mitotic chromosomes: loops-on-a-scaffold models, where DNA loops are attached to a central axial core structure, and hierarchical models, where chromatin is looped or coiled into increasingly thicker fibers. Using polymer simulations, we found that a loops-on-the-scaffold model of metaphase chromosome folding is consistent with the Hi-C data, and previous microscopy and EM measurements, while hierarchical models are not consistent with Hi-C.

Finally, we propose that mitotic chromosome organization consistent with Hi-C and microscopy data can naturally emerge by a combination of two biological processes: initial linear compaction of the chromosome by formation of an array of consecutive loops, and subsequent axial compression of this loop array to form the final compact mitotic chromosome.

Biological membranes are fluid and elastic surfaces that change its shape in the course of key cellular phenomena, such as endocytosis, infection, immune response, division, etc. Curvature also controls the way proteins interact with one another and so it acts as a vital signaling mechanism in the cell. Cell鈥檚 most notable remodelers are BAR proteins, known to both generate membrane curvature and sense complex membrane morphologies. We combine theoretical modeling with experimental biophysical methods to study the driving force underlying the reshaping of biological membranes induced by BAR proteins. In particular, we employ coarse-grained molecular dynamics and continuum-mechanics simulations to study the assembly of proteins on the membrane and elucidate how their association affects membrane鈥檚 large-scale morphology. It also lets us identify a surprising sensitivity of protein-protein attractions on the large-scale mechanics. On the other hand, by using quantitative fluorescence and atomic force microscopies, we investigate the curvature-function relationship of membranes at much larger time and length scales. We study the recruitment of BAR proteins on membrane tubules and the way curvature triggers the formation of rigid protein scaffolds. Finally, we show an unexpected relationship between BAR proteins and molecular motors that drives membrane fission in a newly discovered endocytic mechanism. Our combined theoretical and experimental approach helps explain, at various different resolutions, how BAR proteins regulate membrane dynamics in living cells.

A common metaphor for describing development is a rugged 鈥樷�榚pigenetic landscape鈥欌�� where cell fates are represented as attracting valleys resulting from a complex regulatory network. I previously introduced a framework for explicitly constructing epigenetic landscapes that combines genomic data with techniques from spin-glass physics such as the projection method Hopfield neural network. This model provides predictions for reprogramming

It will be shown that: (a) Shannon information theory, as extended by Jaynes vastly improves discovery of potential risk loci; (b) an allele, rather than a SNP perspective contains more information, and is computationally superior; (c) a rigorous digitalization of genomic symbol data permits application of powerful tools, leading to a disease classifier that numerically quantifies disease likelihood.

The present methodology, which departs substantially from customary practice, proves fruitful in the investigation of Genome Wide Association Studies (GWAS). It extends naturally and successfully to predicting genomic disposition to disease, arising from large collections of weakly contributing loci.

Strong evidence will be advanced that adult onset diabetes (鈥渢ype 2 diabetes鈥�, T2D) is such a candidate disease, and specific genes will be implicated. T2D is characterized by a large pool of potential disease loci. When present in sufficient number in someone then that individual is judged to have diabetes. There are virtually limitless different assemblies of the pool which can be designated as diabetes.

During operant conditioning, animals learn to respond to stimuli with actions that lead to favorable outcomes. Recordings in mammals have uncovered neural signals that link stimulus, action, and outcome, but a limited number of recording sites have hindered the comprehensive discovery of learning signals. Here, we use brain-wide functional imaging in larval zebrafish to screen more than 100,000 neurons while animals learn to terminate a heat stimulus with a directional tail movement. We identify neurons comprising two major classes: class 1 consists of action-selective neurons that encode the direction of heat-evoked tail movements seconds before and after their execution. Class 2 consists of neurons that encode outcome prediction and prediction error. This class includes both positive relief prediction signals that are enhanced by learning, and negative relief prediction signals that are suppressed by learning. These positive and negative relief prediction signals not only have opposing patterns of activity but are also found on opposing sides of the habenula. Strikingly, both outcome-predictive and action-selective signals are correlated with natural variability in learning performance across animals. This study provides the first comprehensive survey of the neural dynamics during learning, suggests that lateralized neuronal activity contributes to operant conditioning, and raises novel hypotheses about the roles of action-selective and outcome-prediction neurons.

We present 2D DNA compartments fabricated in silicon as 鈥榓rtificial cells鈥� capable of metabolism, programmable protein synthesis, and communication. Metabolism is maintained by continuous diffusion of nutrients and products through a thin capillary, connecting protein synthesis in the DNA compartment with the environment. We programmed protein expression cycles and auto-regulated protein levels in separates 鈥榗ells鈥�. We used a genetic switch to program emergent long-range front propagation of gene expression a 1D array of 鈥榗ells鈥� with short-range interactions. The propagation velocity is maximal near a saddle-node bifurcation to a monostable homogenous regime, where propagation breaks down. Near the transition expression onset time exhibits slowing down and strong fluctuations. This demonstrates how on-chip gene circuits as dynamical systems can be programmed to drive spatiotemporal patterns and intercellular communication, cellular variability, and spontaneous symmetric breaking.

Two-dimensional (2D) in vitro culture systems have for a number of years provided a controlled and versatile environment for mechanistic studies of cell adhesion, polarization, and migration, three interrelated cell functions critical to cancer metastasis. However, the organization and functions of the cytoskeleton, focal adhesion proteins, protrusion machinery, and microtubule-based polarization in cells embedded in more physiologically relevant 3D extracellular matrices are qualitatively different from their organization and functions on conventional 2D planar substrates. This talk will describe the implications of the dependence of these fundamental differnces on micro-environmental dimensionality (1D vs. 2D vs.. 3D), how cell micromechanics plays a critical role in promoting local cell invasion, and associated validation in mouse models. We will discuss the implications of this work in cancer metastasis.

How does a nervous system control an animal鈥檚 behavior? We are investigating this question by manipulating and monitoring neural activity of populations of neurons in the brain of a simple transparent organism and correlating the observed neural dynamics with behavior. I will present a suite of optical tools to control and record activity in the nematode C. elegans as it moves, including the first instrument to perform whole-brain calcium imaging with cellular resolution in an awake and unrestrained behaving animal. We have used these techniques to gain insight into the underlying neural mechanisms behind mechanosensation, forward locomotion, and the C. elegans escape response. These measurements are a critical first step for investigating neural coding of behavior, decision-making and the time evolution of internal brain states.

Living organisms exist at many scales of biological organization: prokaryotic organisms, eukaryotic unicellular organisms, multicellular organisms and even eusocial organism such as the societies of ants and bees. But what drove the major evolutionary transitions between levels of biological organization? We investigate this problem using a prokaryotic model: swarming in the bacterium Pseudomonas aeruginosa. Swarming is a collective form of motility whereby billions of bacterial cells come together to move across surfaces. It requires the production of rhamnolipid biosurfactants that wet the surface, but biosurfactant synthesis can be costly to individual cells. We use a combination of mathematical modeling and quantitative experiments to investigate how bacteria make the decision to produce biosurfactants by expressing the genes for rhamnolipid synthesis. Beyond P. aeruginosa we aim to uncover general principles of how individual cells implement social decision-making, a problem with relevance to many systems including the evolution of multicellularity and cancer.

Every lipid membrane can support a transmembrane potential, and membrane voltage affects the dynamics of a huge variety of biological processes. However, this quantity has remained difficult to measure. We engineered a microbial rhodopsin protein to function as a fluorescent reporter of membrane voltage. Archaerhodopsin-derived voltage-indicating proteins enable optical mapping of bioelectric phenomena with unprecedented speed and sensitivity. We are developing molecular tools, transgenic animals, instrumentation, and software to visualize dynamic bioelectric effects in canonical systems (neurons, cardiac cells) and in places where bioelectricity is little studied (bacteria, plants, endothelial cells).

An attractive feature of bottom-up approaches to creating novel materials is that only minimal control of the system is needed at `run time’. For example, in molecular self-assembly, polymer folding, or actuation of mechanical meta-materials, the microscopic interactions are programmed to be so constraining that only the desired behavior is allowed. The implicit assumption is that such single-purpose microscopic interactions, dedicated to only one behavior, are needed in order to get away with minimal dynamical control of the system.

Instead, we find a non-trivial tradeoff between control and multi-functionality by connecting to ideas in theoretical neuroscience (“associative memory鈥欌��). We find that a finite number (“capacity鈥�’) of distinct behaviors can be simultaneously programmed into a system without much incremental need for dynamical control. However, programming any additional behaviors leads to a new regime that requires dramatically finer dynamical control of the system. We illustrate these ideas using models and examples of multi-functional DNA self-assembly, mechanical meta-materials and ribozymes.

We find that Bacillus subtilis bacteria in a growing colony exhibit large (non-thermal) number fluctuations. Motile bacteria within a colony form dynamic clusters defined by nearest neighbors having nearly the same speed and orientation. Surprisingly, the speed and orientational correlations of bacteria within a cluster are scale invariant. Studies of another rod-shaped motile bacterium found commonly in soil, Paenibacillus dendritiformis, reveal that neighboring colonies secrete a toxic protein, Slf, which is not secreted by an isolated colony. Some bacteria within a colony survive Slf but are induced to switch to immotile Slf-resistant cocci. If the cocci encounter sustained favorable conditions, they secrete a signaling molecule that induces a switch back to the rod-shaped form. Genes encoding components of this phenotypic switching pathway are widespread among bacterial species, suggesting that this survival mechanism is not unique to P. dentritiformis.

The origin of multicellularity was one of the most significant innovations in the history of life. Our understanding of the evolutionary processes underlying this transition remains limited, however, mainly because extant multicellular lineages are ancient and most transitional forms have been lost to extinction. We bridge this knowledge gap by evolving novel multicellularity in vivo, using baker鈥檚 yeast and Chlamydomonas reinhardtii as model systems. In this talk I will cover recent work examining: 1) how cells evolve to form multicellular clusters, 2) how these clusters become 鈥楧arwinian individuals鈥� capable of adaptation, 3) how multicellular life cycles that include single-celled genetic bottlenecks arise in evolution (and why this is important), and 4) how nascent multicellular entities evolve to be more complex. Our approach, which allows for the study of macroevolutionary processes over microevolutionary timescales, demonstrates that multicellularity is less evolutionarily constrained than previously thought. If time permits, I will briefly cover ongoing projects in our lab (examining, for example, the origin of multicellular development and the 鈥榬atcheting hypothesis鈥� for multicellular complexity).

The work in our lab focuses on understanding the neural basis of spatial memory and spatial cognition in freely-moving, freely behaving mammals 鈥� employing the echolocating bat as our animal model. I will describe our recent studies, including: (i) Recordings of 3-D place cells, 3-D grid cells, and 3-D head-direction cells in the hippocampal formation of freely-flying bats, using a custom neural telemetry system 鈥� which revealed an elaborate 3-D spatial representation system in the bat鈥檚 brain; and (ii) Absence of theta oscillations in the bat鈥檚 hippocampal formation 鈥� arguing against a central role of theta in spatial cognition. I will also describe our recent studies of spatial memory and navigation of bats in the wild, using micro-GPS devices, which revealed outstanding navigational abilities and provided the first evidence for a large-scale ‘cognitive map’ in a mammal.

In the hippocampal formation, some neurons fire whenever the animal’s location coincides with a point of an imaginary, hexagonal grid that tessellates space. Such neurons, called grid cells, are thought to play an important role in spatial navigation. Achieving a metric representation for space requires more than a single grid cell; indeed, it requires an ensemble of neurons organized into separate modules, with each module representing a different spatial scale. For the population code to yield a metric, I will show that all grid lattices within a module must share the same orientation, which permits decoding through population vectors. The periodic nature of grid cells makes a simple population vector read-out equivalent to maximal likelihood decoding, which is optimal. Furthermore, the read-out can be made robust to distortions in the regular grid pattern near the boundaries of the environment (Stensola et al. 2015) to permit a global grid-cell metric. Computing the homing vector is least error-prone when the ratio of successive grid scales is around 3/2, consistent with the scales found in the mammalian grid code (Stensola et al, 2013). Silencing intermediate-scale modules should cause systematic errors in navigation, while knocking out the module at the smallest scale will only affect navigational precision. In this scheme, read-out neurons behave like goal-vector cells, for which the goal location depends on nonlinear gain fields, similar to the gain fields found in parietal cortex.

I-theory starts from the hypothesis that invariant representations of images are the main computational goal of the ventral stream in visual cortex. Invariant representations can be proved to lead to lower sample complexity in image recognition. We propose a biologically plausible simple-complex cells module (HW module) for computing components of an invariant signature. We extend it to an architecture that in addition to invariance achieve efficient approximation of multidimensional functions by using an extension of additive splines that we call hierarchical additive splines. We show that today鈥檚 Deep Convolutional Networks can be characterized in terms of this theoretical framework.

Performance in sensory perception, time and space perception, decision-making, short-term memory and motor control obeys the Weber (log-scale) law. What neuronal mechanisms can support such a wide dynamic range yet in a well-controlled manner? I will demonstrate that lognormal distributions are fundamental to both structural and functional brain organization. A small minority of fast firing neurons and connections may be responsible for ‘good enough’ brain performance but deployment of the weakly active majority is needed for precision performance. I will illustrate these speculative ideas with concrete examples.

Even though this goal is often professed, one cannot hope to have a “complete” model of any biological system, which would enumerate and quantitatively account for every possible bio-molecular player. Explicit coarse-grained models generalize better, and bring about more understanding. Here I will explore a few examples of coarse-grained models in cellular and neural systems, and will show some unexpected findings that this modeling approach uncovered.

Neural firing rates in the macaque lateral intraparietal (LIP) cortex exhibit gradual “ramping” that is commonly believed to reflect the accumulation of sensory evidence during decision-making. However, ramping that appears in trial-averaged responses does not necessarily indicate that the spike rate ramps on single trials; a ramping average rate could also arise from instantaneous steps that occur at different times on each trial. In this talk, I will describe an approach to this problem based on explicit statistical latent-dynamical models of spike trains. We analyzed LIP spike responses using spike train models with: (1) ramping “accumulation-to-bound” dynamics; and (2) discrete “stepping” or “switching” dynamics. Surprisingly, we found that three quarters of choice-selective neurons in LIP are better explained by a model with stepping dynamics. We show that the stepping model provides an accurate description of LIP spike trains, allows for accurate decoding of decisions, and reveals latent structure that is hidden by conventional stimulus-aligned analyses.

Many small molecules induce general anesthesia in animals ranging from insects to mammals. Although their mechanism of action remains controversial, it is known that a compound鈥檚 efficacy as an anesthetic is strongly predicted by both its hydrophobicity and its potency in inhibiting a wide variety of ligand gated ion channels. Here I will argue that anesthetics interfere with and mimic native regulation of ion channels by their two-dimensional solvent, the plasma membrane, which is tuned close to a liquid-liquid critical point. I will first review the evidence suggesting that plasma membranes are indeed tuned close to such a de-mixing critical point. I will then report on recent experiments where we show that the n-alcohol anesthetics take plasma membrane derived vesicles away from criticality by lowering their critical temperature. I will then discuss our model for understanding anesthetic effects, in which the nearly critical membrane鈥檚 order parameter acts as an allosteric modulator for bound channels. I will show simulation results from the 2D-Ising model that suggest that ion channels could indeed be strongly affected by our measured changes in critical temperatures. Our mechanistic model makes a wide array of predictions that we are beginning to test. I will present our preliminary results showing that several compounds that reverse anesthetic effects on membrane criticality also reverse tadpole anesthesia and functional effects at the channel level. In addition, we predict that channels regulated in this way will have their partitioning into membrane domains altered by function, both in intact cells and in cell-derived vesicles.

I will describe: – the “top-down” approach of understanding brain function by starting from behavior; – probabilistic inference as a framework to qualitatively account for perceptual illusions; – probabilistic inference as a framework to quantitatively account for decision-making behavior in laboratory tasks; – two possible neural implementations of probabilistic inference in a decision-making task.

Recurrent neural networks are an important class of models for explaining neural computations. Recently, there has been progress both in training these networks to perform various tasks, and in relating their activity to that recorded in the brain. Despite this progress, there are many fundamental gaps towards a theory of these networks. Neither the conditions for successful learning, nor the dynamics of trained networks are fully understood. I will present a detailed analysis of very simple tasks as an approach to build a theory of general trained recurrent neural networks.

A holy grail of nano-technology is to create truly complex, multi-component structures by self assembly. Most self-assembly has focused on the creation of `structural complexity’. In my talk, I will discuss `Addressable Complexity’: the creation of structures that contain hundreds or thousands of distinct building blocks that all have to find their place in a 3D structure. Recent experiments have demonstrated the feasibility of making such structures. Simulation and theory yield surprising insights that can inform the design of novel structures and materials.

The self-assembly of cells into 3D structures, such as biological tissues, is important during embryogenesis, morphogenesis and wound healing. Inspired by biological systems, we design and make droplets stabilized by lipid mixtures, which are functionalized with cell-cell adhesion proteins. We discover that lipids phase separate on the droplet surface to create stable and tunable patterns of circular or stripy domains, reminiscent of lipid rafts in cell membranes. These domains carry adhesive proteins, which then drive the specific and reversible binding between droplets to generate large scale structures. For example, we show that these mobile adhesion patches can self-assemble linear chains of droplets into compact structures. These results demonstrate the control of valency, geometry and specificity of droplet bonds to open novel routes to the self-assembly of biocompatible scaffolds for the controlled study of 3D migration of biological cells.

This talk will discuss a non-conventional neural coding task that may apply more broadly to many senses in higher vertebrates. We ask whether and how a non-visual sensory system can focus on an object. We present recent experimental and modeling work that shows how the electric sense can perform such neuronal focussing. This sense is the main one used by weakly electric fish to navigate, locate prey and communicate in the murky waters of their natural habitat. We show that there is a distance at which the Fisher information of a neuron’s response to a looming and receding object is maximized, and that this distance corresponds to a behaviourally relevant one chosen by these animals. Strikingly, this maximum occurs at a bifurcation between tonic firing and bursting. We further discuss how the invariance of this distance to signal attributes can arise, a process that first involves power-law spike frequency adaptation. The talk will also highlight the importance of expanding the classic dual neural encoding of contrast using ON and OFF cells in the context of looming and receding stimuli.

Evolution is simple if adaptive mutations appear one at a time. However, in large microbial populations many mutations arise simultaneously resulting in a complex dynamics of competing variants. I will discuss recent insight into universal properties of such rapidly adapting populations and compare model predictions to whole genome deep sequencing data of HIV-1 populations at many consecutive time points. Genetic diversity data can further be used to infer fitness of individuals in a population sample and predict successful genotypes. We validate these prediction using historical influenza virus sequence data. Successful predictions of the composition of future influenza virus population could guide strain selection for seasonal influenza vaccines.

Wiring in neural circuits ranges from highly stereotyped to apparently random. Using examples of random circuits in flies, fish and mice, I will discuss the logic behind the design and function of random representations in sensory processing, learning and decision making.

Many important behaviors arise from the joint activity of large numbers neurons spread over multiple brain areas. We use a combination of light-sheet microscopy and a virtual reality setup to image activity in almost the entire brain (~100k neurons) of fictively behaving larval zebrafish. Our general approach is to evoke the relevant behavior in the VR setup, record whole-brain activity optically, apply large-scale computational techniques to create functional maps, and use these to guide perturbation experiments and anatomical mapping. We first addressed how structure in spontaneous behavior arises from brain activity. We analyzed the spatiotemporal structure of spontaneous swim trajectories in the absence of any salient sensory cues, in both freely swimming and virtual reality settings. From whole-brain functional maps we identified a relatively small number of neurons in stereotyped locations, which through lesion and stimulation experiments we showed to be causally related to the spontaneous behavior, critical to structuring swim trajectories. Using tracing techniques, we found putative connections between the bias-inducing neurons and premotor neurons that can drive turning behavior, leading to an integrative model of how this circuit controls spatiotemporal patterning of spontaneous locomotion. We also use similar techniques to analyze neural circuitry underlying motor learning and motor memory formation. These studies demonstrate how whole-brain imaging during behavior can localize and map the dynamics of previously unknown functional populations of neurons.

Cognitive flexibility allows us to quickly change our behavior to fit the situation. This is thought to require the dynamic re-mapping of neural circuits in order to select contextually appropriate behaviors. Here I present evidence that changes in synchrony within and between brain regions may underlie such dynamic re-mapping. Leveraging large-scale, multiple-region electrophysiology in non-human primates, I will show different patterns of synchrony in frontal and parietal cortex support different behaviors. In particular, different tasks are encoded by distinct dynamic, synchronous, sub-networks within prefrontal cortex and that these sub-networks organize the spiking activity of single neurons carrying task-relevant information. Such synchronous sub-networks may provide an ideal mechanism for flexibly associating task-relevant information, supporting cognitive flexibility.

In this talk, I am going to review a bit of CRISPR biology, which inspired my colleagues and me to study this fascinating immune system of bacteria. The CRISPR (clustered regularly interspaced short palindromic repeats) mechanism allows bacteria to defend adaptively against phages and other invading genomic material. The CRISPR machinery acquires short genomic sequences (so-called spacers) from the “invaders” and in this way builds up a memory of past infections. With a new encounter of an invading sequence, this memory is accessed, and in a successful outcome the invader is neutralized. I will introduce a population dynamics model where immunity can be both acquired and lost. Two key parameters of the model are the ease of acquisition and the spacer effectiveness in conferring immunity. I will describe the predictions of this model and suggest potential experiments.

A central challenge in neuroscience is to understand connectivity and computations in neural circuits. In the retina, diverse cell types process visual information captured by photoreceptors with exquisite precision and specificity, delivering multiple distinct visual images to the brain. I will present the approach we have taken to understanding retinal computations using large-scale electrical recordings combined with high-resolution visual stimulation. First, I will describe the unique technical approach we have developed for this purpose and the foundational work in the field on which it is based, which will likely be novel to a physics audience. Then, I will show how our large-scale, high-resolution measurements first allowed us to reveal functional connectivity in the retina at cellular resolution and circuit scale a few years ago. Finally, I will show how we have recently extended this approach to decipher one of the elementary nonlinear neural computations that the retina performs on the visual scene.

Complex systems are characterized by an abundance of meta-stable states. To describe such systems statistically, one must understand how states are sampled, a difficult task when thermal equilibrium does not apply. This problem arises in various fields of science, and here I will focus on a simple example, sand. Sand can flow until one jammed configuration (among the exponentially many possible ones) is reached. I will argue that these dynamically-accessible configurations are atypical, implying that in its solid phase sand “remembers” that it was flowing just before it jammed. As a consequence, it is stable, but barely so. I will argue that this marginal stability answers long-standing questions both on the solids and liquid phase of granular materials, and will discuss the applicability of this idea to other systems.

Magnetoencephalography is one of the leading techniques for functional brain studies. So far it has relied on arrays of cryogenic SQUID magnetometers. I will describe a new technique using atomic magnetometers that brings additional capabilities for MEG studies. We have developed a new type of optically-pumped magnetic field sensors that can compete with SQUIDs in many applications. They do not require cryogenic cooling and allow greater flexibility of sensor placement and as well as higher magnetic field sensitivity. Several groups worldwide have been pursuing MEG using these atomic sensors with promising initial results. I will also describe another type of atomic magnetic field sensor that can convert magnetic field magnitude directly to frequency measurements, allowing very high dynamic range and noise rejection. Such sensors recently reached sensitivity competitive with SQUIDs and may allow, for the first time, MEG recordings in magnetically-unshielded environment.

The tree structure is currently the accepted paradigm to represent evolutionary relationships between organisms, species or other taxa. However, horizontal, or reticulate, genomic exchanges are pervasive in nature and confound characterization of phylogenetic trees. Drawing from algebraic topology, we present a unique evolutionary framework that comprehensively captures both clonal and reticulate evolution. We show that whereas clonal evolution can be summarized as a tree, reticulate evolution exhibits nontrivial topology of dimension greater than zero. Our method effectively characterizes clonal evolution, reassortment, and recombination in RNA viruses. Beyond detecting reticulate evolution, we succinctly recapitulate the history of complex genetic exchanges involving more than two parental strains, such as the triple reassortment of H7N9 avian influenza and the formation of circulating HIV-1 recombinants. In addition, we identify recurrent, large-scale patterns of reticulate evolution, including frequent PB2-PB1-PA-NP cosegregation during avian influenza reassortment. Finally, we bound the rate of reticulate events (i.e., 20 reassortments per year in avian influenza). Our method provides an evolutionary perspective that not only captures reticulate events precluding phylogeny, but also indicates the evolutionary scales where phylogenetic inference could be accurate.

It has long been appreciated that the statistical properties of natural stimuli shape neural processing mechanisms in the sensory periphery, but the extent to which such a principle can be formulated for and applied to central processing is unclear. The periphery faces a transmission bottleneck, so efficiency implies compression of signal components with a predictably wider range. Cortex faces a different challenge 鈥� it must use limited samples to make inferences to guide decisions. In this regime, efficient coding predicts the opposite from the periphery: that greater resources are allocated to the signal components with a wider range. To test this hypothesis, we carry out two parallel studies. In one, we measure the joint distribution of local two-, three-, and four-point spatial correlations in an ensemble of natural images. In the other, we measure human perceptual sensitivity to these correlations and their combinations via psychophysical experiments that use synthetic visual textures. We show that psychophysical performance, described by dozens of independent parameters, can be predicted with surprising accuracy from the distribution of spatial correlations found in the natural images. Thus, the efficient coding principle extends beyond the sensory periphery to the central nervous system, where it applies in a very different guise and accounts for the sensitivity to higher-order elements of visual form.

Various functions performed by chromosomes involve long-range communication between DNA sequences that are tens of thousands of bases apart. In this talk I will discuss experiments and theory relating to two distinct modes of long-range communication, chromosome looping and hopping, both in the context of DNA break repair in yeast. In particular, I will show that loss of polymer entropy due to chromosome looping serves as the driving force for homology search during DNA break repair. Also, I will discuss the spread of histone modifications away from the DNA break point in the context of a simple physics models based on a kinase hopping along the chromosome. These examples reveal physical principles of long-range communication in the nucleus, and how they can be tested experimentally.

Cells sense and respond to physical and chemical cues in their environments through networks of interacting proteins–signal transduction systems–that detect and interpret specific inputs, and control the appropriate physiological responses. In bacteria, one of the major modes of signal transduction is mediated by a class of circuits that are composed of two proteins: a sensor kinase and a response regulator. These two-component systems have been found in remarkable numbers within individual organisms and across different bacterial species. They play a central role in regulating basic aspects of microbial physiology and mediate adaptation to diverse environments. I will describe work in which we have explored the organization and properties of these circuits in E. coli.

The core eukaryotic cell cycle is believed to have emerged before the last common ancestor of plants, fungi, and animals. The structure of the G1鈥揝 regulatory network and its dynamic properties are highly similar in budding yeast and mammals despite lack of sequence homology between many G1/S regulators (cyclins, transcription factors, and inhibitors). This is all the more striking because plant and animal G1/S regulators (Cyclin D, E2F/DP, Rb) have much higher sequence homology even though fungi and animals are more closely related.

To investigate cell cycle evolution, we constructed a eukaryotic phylogeny of the G1/S cyclin, cyclin-dependent kinase, inhibitor, and transcription factor super-families. We show that the G1/S regulatory network present in plants and animals was present in the last common eukaryotic ancestor, and that a fungal ancestor evolved novel G1/S cyclins (CLN), inhibitor (Whi5), and transcription factors (SBF/MBF). The fungal G1/S cyclins (CLN) co-cluster with B-type cyclins found in plants and animals. Basal Fungi are characterized by the emergence of a novel Whi5 inhibitor and novel SBF/MBF transcription factors followed by the eventual loss of the ancestral Rb inhibitor and E2F/DP transcription factors. Strikingly, SBF/MBF DNA-binding domain has sequence homology to the KilA-N domain, which suggests a viral origin for this fungal-specific family of transcription factors. It appears that these new regulators were able to acquire, maintain, and eventually usurp the function and dynamics of the original G1-S regulatory network following the formation of a hybrid network. The emergence of this hybrid SBF-E2F cell cycle coincides with the emergence of Fungi.

The identifiability problem is critical problem in the inference of biological networks from genomics data. Thus, even very large observational data-sets are not sufficient for discriminating between a very large number of equally likely network models. Even as the size and quality of expression, proteomics and phospho-proteomics data-sets increases the quality and completeness of large scale regulatory network models derived from these data-sets might not increase. One way to approach this problem is by collecting data-sets that generate priors on network structure (such as ChIP-seq, ATAC-seq, and others). I will discuss new computational methods for using priors in network structure to learn dynamic regulatory network models. These methods use priors on network structure in two ways: 1) to estimate latent activities of transcription and chromatin remodeling factors and 2) to constrain the model selection step of our method for learning dynamic regulatory network models, the Inferelator. Examples application to B. subtilis (with Patrick Eichenberger) and the mouse immune system (with Dan Littman) will be discussed. Lastly, implications for optimal experimental design of large genomics data-collection efforts will be discussed.

Cell assemblies are thought to be the units of information representation in the brain, yet their detection from experimental data is arduous. We propose to infer the effective network structure from simultaneously recorded neurons in prefrontal cortex using an inverse Ising model. We define cell assemblies as the co-activated neurons in the dynamics of the resulting abstract neural network. Different dynamical regimes, as may be observed across wakefulness and sleep, can be reproduced by changing a global input parameter, providing access to those rare activity configurations that are crucial for cell assembly replay. The identified assemblies strongly co-activate during wakeful experience and are found to replay during subsequent sleep, in correspondence to the potentiation of the inferred network. Across sessions, a variety of different network scenarios is observed, providing insight in cell assembly formation and replay.

Intelligent systems, whether natural or artificial, must act in a world that is highly unpredictable. It is intuitively clear, that an optimal approach to decision making or planning under such circumstances requires to take these uncertainties into account. However, the optimal control solution is intractable to compute in general and in addition is hard to represent, due the non-trivial state dependence of the optimal control. This has prevented large scale application of stochastic optimal control theory sofar. The path integral control theory describes a class of control problems whose solution can be computed as an inference computation in a probability model. This link between control and inference may provide an opportunity to treat learning and acting within the same theoretical framework. In this talk, I will review the theory and discuss the relation to statistical physics. I will then discuss how Monte Carlo importance sampling provides an elegant approach to learning motor control.

The fish fauna of the Southern Ocean is unlike that of any other set of near-shore habitats on Earth as it is completely dominated by a lineage of closely related species: the notothenioids. Over the course of approximately 35 million years, notothenioid fishes have acclimated to the changing environmental conditions of the Southern Ocean and now have a suite of adaptations to avoid freezing in the subzero waters surrounding Antarctica. In this talk, I explore the idea that the ecological and morphological diversity exhibited in Antarctic notothenioids is facilitated by the “key innovation” of anti-freeze glyco proteins (AFGPs). My analyses utilize hypotheses of phylogenetic relationships and estimates of divergence times among notothenioids to reconstruct the history of their diversification, and place key events in the evolutionary history of notothenioids with our understanding of the paleoclimate of Antarctica.

We want to make a 鈥渘on-biological鈥� system which can self-replicate. The idea is to design particles with specific and reversible and irreversible interactions, introduce seed motifs, and cycle the system in such a way that a copy is made. Repeating the cycle would double the number of offspring in each generation leading to exponential growth. Using the chemistry of DNA either on colloids or on DNA tiles made from three linked DNA helices, BTX, or on DNA origamis makes the specific recognition part easy. In the case of DNA BTX tiles we have in fact replicated the seed at least to the third generation. In more recent work we have used DNA origamis that can be permanently crosslinked using a UV activated molecule, CNV, on DNA sticky ends. This system needs no external intervention other than cycling temperature and light, mimicking daily cycles on earth. The system exponentially grows and through 24 doubling cycles has produced offspring multiplying the initial seed more than 7,500,000 times. We even have initiated an elementary form of evolution and shown selection of one species over another. The process should be adaptable to assemble and replicate micro-nano constructs and devices.

In the first part of this talk I will review multi species competition by resources, which defines a niche axis. I will discuss the emergence of a lumpy pattern for species abundances along this niche axis, a problem that was called “the paradox of the clumps”, and how this problem can be solved. I will also present some important results concerning the total abundance and diversity of ecological communities. In the second part of this talk I will propose a way to extend niche competition theory by including interactions different from mutual competition, like mutualism or symbiosis and parasitism, via game theory. I will analyze the effects on total abundance and diversity of introducing cooperative interspecific interactions for different popular games (prisoner’s dilemma and chicken).

To efficiently catalyze multi-step biochemical reaction pathways, cells have optimized the synergistic action of a multitude of enzymes. They not only control the concentrations and activities of enzymes, but often also coordinate enzymes from the same biochemical reaction pathway by arranging them in self-assembled multi-enzyme complexes. Such complexes are the basis of 鈥渃hanneling鈥� mechanisms, where intermediate products in multi-step reactions are effectively passed from one enzyme to the next. These same principles can be applied in artificial nano-scale systems, with the ultimate goal to control and optimize biochemical reactions at will, e.g. for the production of medical substances or renewable energy sources. While enzymatic activity has been studied for over a century, quantitative experiments were limited to the bulk level until the recent advent of single-molecule enzymology techniques, which permit the quantitative characterization of the stochastic reaction dynamics of enzymes. Single-molecule experiments with enzymes have also spurred the theoretical analysis of stochastic enzyme dynamics, but the theoretical description and analysis of enzyme complexes is only beginning to develop. I will describe some recent progress in this direction.

During embryonic development, an entire organism is generated from a single cell. Genetics and biochemistry have identified developmental signaling pathways, however, how embryonic patterns emerge in space and time remains more obscure. I will discuss our work utilizing micropatterning approaches to generate early developmental patterns starting from pluripotent human embryonic stem cells. Our work reveals how cell-to-cell communication generates self-organized patterns and establishes micropatterned stem cell culture as a promising system for addressing fundamental questions in developmental biology quantitatively.

For important biological functions such as wound healing, embryonic development, and cancer tumorogenesis, cells must initially rearrange and move over relatively large distances, like a liquid. Subsequently, these same tissues must undergo buckling and support shear stresses, like a solid. Our work suggests that biological tissues can accommodate these disparate requirements because the tissues are close to glass or jamming transition. This is important because most existing studies of disease focus on single-cell motility, but in glassy tissues the dominant contributions to cell migration come from collective effects and constraints imposed by neighbors. I will discuss a new theoretical framework for predicting rates of cell migration in epithelial (skin) cell layers, and explain how similar models predict surface tension in tissues and cell shape changes that generate left-right asymmetry in embryos. I will also discuss our current work to predict how cancer cells migrate through dense tissues and understand how active cell processes (such as cell polarization) alter the physics of glasses.

Irreversibility is a fundamental aspect of the evolution of natural systems, and quantifying its manifestations is a challenge in any attempt to describe non-equilibrium systems. In the case of fluid turbulence, an emblematic example of a system very far from equilibrium, the motion of a single fluid particle provides a clear manifestation of time irreversibility. Fluid particles tend to lose kinetic energy faster than they gain it. This is best seen by the presence of rare “flight-crash” events, where fast moving particles suddenly decelerate into a region where fluid motion is slow. Remarkably, the statistical signature of these events establishes a quantitative relation between the degree of irreversibility and turbulence intensity.

Locomotion of animals and robots is well studied in environments composed of hard ground, air, and water; less is known about principles of movement in granular materials, collections of a-thermal particles that display solid, fluid, and gas-like features. I will discuss two examples which illustrate the sensitivity of granular locomotion to changes in movement strategies and substrate properties (like compaction and incline angle). Specifically, I will discuss studies of a limbless snake-like robot (in collaboration with Prof. Howie Choset, CMU) and a flipper based crawling robot. We use these robots as physical models of biological organisms; I will discuss how the robot studies give insight into neuromechanical control principles governing high performance of the animals, in particular the ability of the organisms to control ground reaction forces to remain below granular yield forces.

For more info click here.

The ecology and dynamics of many microbial systems are shaped by how microbes respond to evolving nutrient gradients and microenvironments. Here we show how the response of a sulfur-oxidizing microbe to changing oxygen gradients cause cells to organize into large-scale fronts. We show that these dynamics occur in two steps. First, chemotactic cells moving up the oxygen gradient form a front that propagates with constant velocity. We then show, through observation and mathematical analysis, that this front becomes unstable to changes in cell density. Random perturbations in cell density create oxygen gradients, which lead to the formation of millimeter-scale fluid flows. We argue that this flow results from a nonlinear instability excited by stochastic fluctuations in the density of cells. These results provide a mathematically tractable example of how collective phenomena in ecological systems can arise from the individual response of cells to a shared resource.

Birdsong is a well-established model for neuroscientists interested in how a brain reconfigures itself through learning, in order to achieve a complex behavior. Yet, unveiling how the different brain areas contribute to the generation of the actual motor patterns involved in the generation of song has proven a difficult issue. In this talk I will describe recent advances in the description of the dynamics presented by the avian vocal organ, particularly how reasonably simple motor instructions can drive it to generate sounds with complex acoustic features.

Systems and circuit-level neuroscience has entered a golden age: with modern fast computers, machine learning methods, and large-scale multineuronal recording and high-resolution imaging techniques, we can analyze neural activity at scales that were impossible even five years ago. One can now argue that the major bottlenecks in systems neuroscience no longer lie just in collecting data from large neural populations, but rather in understanding this data. I’ll discuss several neuroscience problems where modern statistical approaches can have a major impact on our understanding of how networks of neurons are connected and how these networks encode and process information.

Invariant object recognition is a fundamental question of sensory neuroscience. Olfaction, being arguably the most genetically tractable and anatomically compact sensory system, presents a great opportunity to approach this basic question. In spite of accumulated knowledge of neuronal anatomical organization and information processing in olfaction, computational principles and the neuronal basis of odor recognition still remain unsolved problems. Here I present a model of concentration invariant odor recognition based on the temporal ranking of receptor responses. I鈥檒l present strong experimental evidence towards this new model and discuss the model prediction for broader questions in olfaction.

Abstract: For important biological functions such as wound healing, embryonic development, and cancer tumorogenesis, cells must initially rearrange and move over relatively large distances, like a liquid. Subsequently, these same tissues must undergo buckling and support shear stresses, like a solid. Our work suggests that biological tissues can accommodate these disparate requirements because the tissues are close to glass or jamming transition. While recent self propelled particle models generically predict a glass/jamming transition that is driven by packing density 蠁 and happens at some critical 蠁c < 1, many confluent biological tissues appear to undergo a jamming transition at a constant density (蠁 = 1). I will discuss a new theoretical framework for predicting energy barriers and rates of cell migration in 2D tissue monolayers, and show that this model predicts a novel type of jamming transition, which takes place at constant 蠁 = 1 and depends only on single cell properties such as cell-cell adhesion, cortical tension and cell elasticity. I will discuss the implications of these ideas for understanding cancer tumor boundaries and metastasis and pattern formation during development.

To make sense of the sensory clutter in the environment, animals must segregate objects of interest from ever varying backgrounds 鈥� for example, picking out a single speaker鈥檚 voice at a noisy cocktail party. In our group, we are investigating the algorithmic and neural bases of such figure-background separation in the olfactory system in mice. We have found that mice can be trained to recognize target odorants embedded in unpredictable and variable background mixtures with high degree of success (>95%). Performance drops with increasing number of background odors, and was well predicted by the extent of overlap between the neural representations of the target and the background odors. We found that a linear classifier (a logistic regression with sparse weight prior) can be trained to detect the target odorant independent of background distractors, using only a small subset of the neural responses. Our study indicates that the olfactory system has powerful analytic abilities that are constrained by the limits of combinatorial neural representation of odorants at the level of the olfactory receptors.

I will summarize an top-down quantitative approach developed by our lab in recent years to elucidate some classic puzzles in bacterial metabolism and cell growth, including catabolite repression, hierarchical substrate utilization, and overflow metabolism (the bacterial Warburg effect).

An incredible gulf separates theoretical models of synapses, often described solely by a single scalar value denoting the size of a postsynaptic potential, from the immense complexity of molecular signaling pathways underlying real synapses. To understand the functional contribution of such molecular complexity to learning and memory, it is essential to expand our theoretical conception of a synapse from a single scalar to an entire dynamical system with many internal molecular functional states. Moreover, theoretical considerations alone demand such an expansion; network models with scalar synapses assuming finite numbers of distinguishable synaptic strengths have strikingly limited memory capacity. This raises the fundamental question, how does synaptic complexity give rise to memory? To address this, we develop new mathematical theorems elucidating the relationship between the structural organization and memory properties of complex synapses that are themselves molecular networks. Moreover, in proving such theorems, we uncover a framework, based on first passage time theory, to impose an order on the internal states of complex synaptic models, thereby simplifying the relationship between synaptic structure and function.

We also apply our theories to model the time course of learning gain changes in the rodent vestibular oculomotor reflex, both in wildtype mice, and knockout mice in which cerebellar long term depression is enhanced; our results indicate that synaptic complexity is necessary to explain diverse behavioral learning curves arising from interactions of prior experience and enhanced LTD.

Many organisms exhibit a systematic left/right or screw assymmintry in their macroscopic form — not just the well-known case of vertebrate viscera. I will concentrate on two stories: “spiral cleavage” in snails, and plant roots (as deternined by microtubulue arrays at the cell level). All must is ultimitely come from the chirality of our proteins, but how does the information get up to the macro scale? I argue that the answer requires physics concepts — forces, motions, and symmetry breaking. Dynamics far from equilibrium is called for, which is supplied in all cases by the cytoskeleton.

Reference article: “Possible origins of macroscopic left-right asymmetry in organisms”, J. Stat. Phys. 148, 740-774 (2012).

Bacterial biofilms are organized communities of cells living in association with surfaces. The hallmark of biofilm formation is a well defined spatio-temporal pattern of gene expression, leading to differentiation and a complex morphology. While this process resembles the development of a multicellular organism, biofilms are only transiently multicellular. More importantly the functions associated to the biofilm phenotype are largely unknown. In this talk I will first discuss aspects of biofilm physiology connected to biofilm expansion on surfaces in the soil bacterium Bacillus subtilis. I will then describe a framework (in progress) to probe the origin of the observed gene expression patterns, a first step to understand the basis for cell decision-making in a biofilm.`

Our concept of a stable genome is changing from one in which the DNA sequence is passed faithfully from generation to generation to another in which genomes are plastic and responsive to environmental changes. Growing evidence shows that environmental stresses induce mechanisms of genomic instability in bacteria, yeast, and human cells, generating occasional fitter mutants and potentially accelerating evolution and disease. The emerging molecular mechanisms of stress-inducible mutagenesis vary but share telling common components that underscore two common themes of the non-randomness of mutation: (1) regulation of mutagenesis in time by cellular stress responses, which promote mutations specifically when cells are poorly adapted to their environments鈥攊.e., when they are stressed; (2) restriction of mutagenesis in genomic space causing mutation hotspots, clusters and showers. Mutational hotspotting may target specific genomic regions and allow local concerted evolution (adaptive evolution requiring multiple mutations). This talk will focus on a molecular mechanism of stress-induced mutation in E. coli and note its parallels in other organisms including human cancer. The mechanism is a stress-response-orchestrated switch to error-prone repair of DNA breaks. We consider its regulation by stress responses, demonstrate its formation of mutation hotspots near DNA breaks, and report our discovery of a large gene network that underlies mutagenic repair of DNA breaks, more than half of which functions in stress sensing and signaling. The data show the importance of stress-response control and also that biological functions of large fractions of networks can be understood when molecular mechanisms are considered and functional studies employed. Regulation of mutagenesis in time and genomic space is widespread in many organisms and circumstances. Such mechanisms may fuel biological evolution and genetic disease, including pathogen-host adaptation and drug resistance and tumor and other disease progression and resistance mechanisms, much of which occurs under stress, driven by mutations.

Our lab focuses on mechanical studies of fundamental biological processes in biology: transcription and replication. These are highly dynamic processes that are carried out by molecular motors that translocate along, and rotate around, DNA. We are interested in forces and torques generated by these motor proteins and how their mechanics regulates transcription and replication. To directly measure these processes at the single molecule level, we develop state-of-the-art (and often one-of-a-kind) instruments and novel techniques. Our recent efforts hold promise to make these precision measurements high throughput on a nanophotonic platform.

In different biological processes, cells move in a coordinated way. Several experiments have quantitatively investigated this phenomenon. I will describe experimental results obtained by the team of P Silberzan (Institut Curie, Paris) as well as a simple model of interacting persistent random walkers that we have developed to phenomenologically describe collective cell motion. The model helps to explain the observed dynamics of a confined cell assembly which displays stochastic reversals of global rotational motion and pulsatile collective modes.

It has long been proposed that turgor pressure plays an essential role during bacterial growth by driving mechanical expansion of the cell wall. This hypothesis is based on analogy to plant cells, for which this mechanism has been established, and on experiments in which the growth rate of bacterial cultures was observed to decrease as the osmolarity of the growth medium was increased. To distinguish the effect of turgor pressure from pressure-independent effects that osmolarity might have on cell growth, we monitored the elongation of single Escherichia coli cells while rapidly changing the osmolarity of their media. By plasmolyzing cells, we found that cell-wall elastic strain did not scale with growth rate, suggesting that pressure does not drive cell-wall expansion. Furthermore, in response to hyper- and hypoosmotic shock, E. coli cells resumed their pre-shock growth rate and relaxed to their steady-state rate after several minutes, demonstrating that osmolarity modulates growth rate slowly, independently of pressure. Oscillatory hyperosmotic shock revealed that while plasmolysis slowed cell elongation, the cells nevertheless 鈥渟tored鈥� growth such that once turgor was re-established the cells elongated to the length that they would have attained had they never been plasmolyzed. In contrast, Bacillus subtilis cells exhibit highly regular growth oscillations in response to hypoosmotic shock that are dependent on peptidoglycan synthesis. The period of these oscillations scales linearly with the magnitude of the shock. By applying a simple mathematical theory to these data, we show that growth oscillations are initiated by mechanical-strain-induced growth arrest. This demonstrates that B. subtilis has developed an elegant system by which turgor pressure both up- and down-regulates the final steps of cell growth.

Microorganisms in all kingdoms of life face a challenge of regulating the size and shape of their cells, control of which is essential for their viability. For decades, a popular hypothesis has been that cells can measure their absolute size, and that reaching a critical size triggers the division process. This would imply that a cell that was born smaller than average will not be smaller than average when it divides 鈥� in contrast to experiments showing that such correlations exist, and that size is partly 鈥渋nherited鈥�. I will present a biophysical model that sheds new light on this problem, showing that cell does not need to know its absolute size to regulate size robustly, quantitatively explaining the experimentally measured correlations in both E. coli and budding yeast, and predicting that average cell size should depend exponentially on the growth rate.

Some biological signalling pathways have to filter out useful information from a vast excess of spurious inputs. An instance of this problem is immune recognition, where T-cells interact mostly with self ligands but should trigger response only when interacting with not-self ligands. Using computational evolution, we discovered a generic functional network (that we called “adaptive sorting”) able to perform such discrimination. An adaptive sorting scheme can be elaborated to account for many puzzling features of early immune detection. Well-known experimental properties such as ligand antagonism or non-monotonicity of response curve appear as natural consequences of sharp detection, and thus might generically occur in other chemodetection systems.