MBI Videos

CTW: Modeling and Inference from Single Molecule to Cells

  • video photo
    Ashok Prasad
    Cells from different tissues typically look quite different from each other even when cultured on plastic or glass slides under identical conditions. This leads us to formulate the hypothesis that cell shape is a function of the cytoskeletal properties of those cells, and begs the question as to what information changes in cell shape carry. This question becomes all the more interesting for cancer, since invasive cancer cells are reported to have altered mechanical properties compared to non-invasive cancer cells. Inspired by this reasoning we study shape characteristics of paired osteosarcoma cell lines, each consisting of a less metastatic parental line and a more metastatic line, derived from the former by in vivo selection. Statistical analysis shows that shape characteristics of the metastatic cell lines are partly overlapping but on average distinguishable from the parental line. Significantly the shape changes fall into two categories, with three paired cell lines displaying a more mesenchymal-like morphology, while the fourth displaying a change towards a more rounded morphology. A neural network algorithm could distinguish between samples of the less metastatic cells from the more metastatic cells with near perfect accuracy. Thus subtle changes in shape carry information about the genetic changes that lead to invasiveness and metastasis of osteosarcoma cancer cells. The next challenge is to link these changes in shape with changes in mechanical cytoskeletal parameters. I will briefly discuss ongoing experiments to infer these cellular mechanical properties by studying internal fluctuations of organelles.
  • video photo
    Raghuveer Parthasarathy
    In each of our digestive tracts, trillions of microbes representing hundreds of species colonize local environments, reproduce, and compete with one another. The resulting ecosystems influence many aspects their host’s development and health. Little is known about how gut microbial communities vary in space and time: how they grow, fluctuate, and respond to perturbations. To address this, we apply light sheet fluorescence microscopy to a model system that combines a realistic /in vivo/ environment with a high degree of experimental control: larval zebrafish with defined subsets of commensal bacterial species. Light sheet microscopy enables three-dimensional imaging with high resolution over the entire intestine, providing visualizations that would be difficult or impossible to achieve otherwise. Quantitative analysis of image data enables measurement of bacterial abundances and distributions and the construction of realistic models of population dynamics. I will describe this approach and focus especially on recent experiments in which a colonizing bacterial species is challenged by the invasion of a second species. Imaging reveals dramatic population collapses driven by peristaltic activity, which differentially affects the two species due to their distinct community architectures. Our findings demonstrate that stochastic perturbations and the physical properties of the host environment can play major roles in determining population dynamics in the vertebrate gut.
  • video photo
    Albert Siryaporn
    Bacteria encounter a variety of mechanical forces during the course of growth and infection. Our lab explores how bacteria detect and respond to forces generated by fluid flow, which is common in many bacterial habitats and host organisms. We find that surprisingly, the bacterium Pseudomonas aeruginosa moves upstream, in the opposite direction, of flow. Cells attach to surfaces at the liquid-surface interface and are oriented upstream by the force of the flow. We detail this mechanism of upstream migration through single-cell measurements and modeling and explore the consequences of this behavior at the multi-cellular level. In particular, we explore how bacteria colonize complex flow networks found in the vasculature of host organisms. Our results show that the interplay between flow and bacterial physiology plays a critical role in determining colonization, competition between different bacterial species, and the dispersal of bacteria. Importantly, our model establishes a foundation for understanding how bacteria grow and spread during pathogenesis.
  • video photo
    Brian Munsky
    Stochastic fluctuations can cause genetically identical cells to exhibit wildly different behaviors. Often labeled "noise," these fluctuations are frequently considered a nuisance that compromises cellular responses, complicates modeling, makes predictive understanding and control all but impossible. However, if we examine cellular fluctuations more closely and match them to discrete stochastic analyses, we discover virtually untapped, yet powerful sources of information and opportunities. In this talk, I will present our collaborative endeavors to integrate single-cell experiments with precise stochastic analyses to gain new insight and quantitatively predictive understanding for Mitogen Activated Protein Kinase (MAPK) signal-activated gene regulation. I will explain how we experimentally quantify transcription dynamics at high temporal (1-minute) and spatial (1-molecule) resolutions; how we use precise computational analyses to model this data and efficiently infer biological mechanisms and parameters; how we predict and evaluate the extent to which model constraints (i.e., data) and uncertainty (i.e., model complexity) contribute to our understanding. I will finish with the discussion of new opportunities in which noise analysis not only helps us to better understand gene regulation phenomena, but where it actually introduces new opportunities to more precisely control these phenomena.
  • video photo
    Kingshuk Ghosh
    Protein sequence encodes complex network of interactions and it is difficult to decipher simple rules in protein science. In spite of this challenge, approximate and semi-empirical rules can be found to describe biophysical properties of different proteins. Using statistical mechanical models tested against multitude of data, our goal is to unravel such universal features of proteins. Our next goal is to extend these transferrable laws in a high throughput manner to model the entire collection of proteins inside an organism, called the proteome. The application at the proteome level allows us to bridge the gap between molecular biophysics and cellular physics and provides us evolutionary insights. With this approach we will try to address some questions of broad interest: i) Why are cells so sensitive to temperature? ii) How do thermophilic proteins (derived from organisms that thrive at high temperature) withstand high temperatures compared to their mesophilic (organisms that live at room temperature) counterparts? iii) What is the evolutionary implication of distribution of different rate processes in a cell and how are they optimized? iv) How do salts slow down cell growth?
  • video photo
    Lisa Lapidus
    An important aspect of protein folding is understanding how folding competes with aggregation, which leads to diseases such as Parkinson’s and Alzheimer’s. The complexity and dynamics of unfolded protein ensembles may be the ultimate speed limit of folding and play a crucial role in aggregation. In my lab over the past several years we have investigated the reconfiguration dynamics unfolded proteins by measuring the rate of intramolecular diffusion, the rate one part of the chain diffuses relative to another. We have measured diffusion coefficients ranging over three orders of magnitude and observed that aggregation-prone sequences tend to fall in the middle of this range. In this talk, I shall present our experiments on alpha-synuclein, the Alzheimer’s peptide and various prion sequences. We correlated intramolecular diffusion of the disordered protein with solution conditions that promote aggregation. Finally, we have begun measurements on small molecule aggregation inhibitors and found that some can prevent aggregation by shifting intramolecular diffusion out of the dangerous middle range.
  • video photo
    Steve Presse
    Abstract not submitted
  • video photo
    Phil Nelson
    It is sometimes said that "Our eyes can see single photons." This article begins by finding a more precise version of that claim and reviewing evidence gathered for it up to around 1985 in two distinct realms, those of human psychophysics and single-cell physiology. Finding a single framework that accommodates both kinds of result is then a nontrivial challenge, and one that sets severe quantitative constraints on any model of dim-light visual processing. I'll present a new model that accomplishes this task, and compare it to recent experiments.
  • video photo
    John Fricks
    In cellular systems, Brownian forces play a dominant role in the movement of small (and not so small) particles such as vesicles, organelles, etc. However, proteins and other macromolecules bind to one another, altering the underlying Brownian dynamics. In this talk, classical approaches in the biophysical literature to time series which switch between bound and unbound states will be presented, and an alternative approach using stochastic expectation-maximization algorithm (EM) combined with particle filters will be proposed. As an example system, molecular motors, such as kinesin, switch between weakly and strongly bound states, as well as directed transport. I will discuss the analysis of such a system along with the ramifications for multi-motor-cargo complexes found in living cells.
  • video photo
    Nils Walter
    Nature employs nanoscale machines that self-assemble into dynamic structures of complex architecture and functionality. Single molecule fluorescence microscopy offers a non-invasive tool to probe and ultimately dissect and control these nanoassemblies in real-time, often with the aid of computational tools to interpret and complement the experimental data. In particular, single molecule fluorescence resonance energy transfer (smFRET) allows us to measure biologically relevant distances and changes thereof at the 2-8 nm scale, whereas complementary super-resolution localization techniques based on Gaussian fitting of imaged point spread functions measure distances in the 10 nm and longer range. In this talk, I will describe how we have developed Single Molecule Cluster Analysis (SiMCAn) based on a vast smFRET dataset to dissect the complex conformational dynamics of a pre-mRNA as it is spliced by the spliceosomal processing machinery. In addition, I will demonstrate how a single molecule systems biology can be implemented that feeds super-resolved single particle tracking data into lattice-based Monte-Carlo simulations to generate novel hypotheses on the gene regulation of mRNAs by microRNAs during RNA silencing.
  • video photo
    Douglas Shepherd
    Abstract not submitted
  • video photo
    Anastasios Matzavinos
    In this talk, we discuss dissipative particle dynamics (DPD) simulations of the dispersion of DNA molecules conveyed by a pressure-driven fluid flow across a periodic array of entropic barriers. We compare our simulations with nanofluidic experiments, which show the DNA to transition between various types of behaviors as the pressure is increased, and discuss physical insights afforded by the ability of the DPD method to explicitly model flows in the system. Finally, we present anomalous diffusion phenomena that emerge in both experiment and simulation, and we illustrate similarities between this system and Brownian motion in a tilted periodic potential. This is a joint work with Clark Bowman, Daniel Kim, and Derek Stein.
  • video photo
    Vasileios Maroulas
    We focus on the biological problem of tracking organelles as they move through cells. In the past, most intracellular movements were recorded manually, however, the results are too incomplete to capture the full complexity of organelle motions. An automated tracking algorithm promises to provide a complete analysis of noisy microscopy data. In this talk, we will get exposed to two such algorithms. The first algorithmic implementation adopts statistical techniques from a Bayesian random set point of view. Instead of considering each individual organelle, we examine a random set whose members are the organelle states and we establish a Bayesian filtering algorithm involving such set-states. The second method combines stochastic filtering estimates of the displacement field with the topological properties on the space of trajectories. Both methods were used to analyze real data.
  • video photo
    Jayajit Das
    Individual isogenic immune cells respond to identical stimuli with unique signaling kinetics. Shapes of kinetic trajectories describing time evolution of abundances of multiple signaling molecules in single cells contain key information regarding signaling mechanisms. However, it can be challenging to measure many signaling reporters simultaneously in single cell in experiments. Flow and mass cytometry experiments can assay a large number of proteins (4 to100) but individual cells are not tracked in these experiments, therefore, such measurements only provide a statistical description of the signaling kinetics, e.g., mean abundances, or, covariance between protein abundances. Is there a way to reconstruct signaling trajectories, even approximately, in individual cells using cytometry data? We address this question affirmatively by using a novel method based on identification of a dynamical invariant pertaining to chemical reaction networks. We validate our method in data obtained from in silico networks and published single cell experiments. We apply our trajectory reconstruction method to analyze mass cytometry data for NaturalKiller (NK) cells to decipher mechanisms underlying NK cell cytotoxic responses.
  • video photo
    Andrew Mugler
    Single cells sense their environment with remarkable precision. At the same time, cells have evolved diverse mechanisms for communicating. How are sensing and communication related? I will describe recent theoretical and experimental results in which this question is explored in several contexts, including gradient detection by connected epithelial cells, and collective invasion of breast cancer cells. I will show how communication allows cells to perform qualitatively new behaviors that single cells cannot perform alone. Moreover, I will demonstrate that minimal mathematical modeling yields fundamental limits to the precision of sensing, and that these limits are critically altered by cell-to-cell communication. This work extends the study of cellular sensing and information processing to collective ensembles.
  • video photo
    Kevin Janes
    Regulated changes in gene expression underlie many biological processes, but globally profiling cell-to-cell variations in transcriptional regulation is problematic when measuring single cells. We have developed an approach, called stochastic profiling, that applies probability theory to transcriptome-wide measurements of small pools of cells to identify single-cell regulatory heterogeneities (Nat Methods 7:311-7 [2010]). I will talk about work in progress that applies stochastic profiling as a tool for uncovering the mechanistic basis of phenotypes that are incompletely penetrant. Regulatory-state frequencies are matched to downstream phenotype frequencies to converge upon a tractable set of candidate states worth of follow-up experimentation. Using the ErbB2 oncoprotein as a model trigger for an incompletely penetrant phenotype, we identify a handful of surprising candidates that significantly affect penetrance when perturbed. Stochastic profiling remains the only method compatible with cells microdissected in situ and thereby opens exciting opportunities in the areas of tissue morphogenesis and cancer.

View Videos By