Program

Speaker: Mason A. Porter (UCLA)

Title: Opinion Dynamics and Spreading Processes on Networks

Abstract: People interact with each other in social and communication networks, which affect the processes that occur on them. In this talk, I will give an introduction to dynamical processes on networks. I will focus my discussion on opinion dynamics, and I will also discuss coupled opinion and disease dynamics on networks. Time-permitting, I may also briefly discuss a model of COVID-19 that centers on disabled people and their caregivers.

Speaker: Krešimir Josić (University of Houston)

Title: Spatiotemporal dynamics of synthetic consortia

Abstract: Modeling is essential for the rational design of genetic circuits with desired properties. I will review several examples where mathematical models have been central to the development and understanding of the dynamic of synthetic organisms. I will how oscillations, and other spatiotemporal patterns can arise in consortia of cells that individually exhibit bistable dynamics. I will show how simplified mathematical models can help us understand how order emerges in these system, how robust oscillations and other patterns can arise, and how they are maintained.

Speaker: Jae Kyoung Kim (KAIST / IBS)

Title: Analysis of dynamic data: from molecule to behavior

Abstract: Due to the advances in experimental technique and wearable devices, dynamic time-series data of molecules to human behavior can be easily measured. In this talk, I will talk about how to use mathematical modeling and theory to infer information underlying time-series data. Specifically, I will talk about the inference of regulatory interactions among molecules from time-series. Furthermore, I will also discuss about the inference of sleep pressure, circadian rhythms, and alertness of human from timeseries data measured by smart watch.

Speaker: Ankit Gupta (ETH Zürich)

Title: Computational methods for analysis and identification of stochastic reaction networks

Abstract: Single-cell studies have revealed that genetically identical cells, grown under similar conditions, can exhibit a significant degree of heterogeneity. The source of this heterogeneity is the randomness in the timing of reactions that constitute intracellular networks and pathways. Stochastic models of reaction networks are extensively used for capturing this intrinsic randomness and quantifying its effects. However, identification of such models with biological experimental data is extremely challenging. The aim of this talk to present an overview of the computational methods that we have been developing for analysing stochastic reaction models and calibrating them with biological data. Specifically, I will present methods for parametric sensitivity analysis, estimating the frequency spectrum of single-cell trajectories, and solving the underlying Chemical Master Equation (CME) and the associated stochastic filtering problem.

Speaker: Dae Wook Kim (University of Michigan)

Title: Efficient assessment of real-world dynamics of circadian rhythms

Abstract: Recent laboratory studies have made remarkable progress in understanding circadian physiology and its clinical implications. However, to apply these findings in real-world clinical settings, it's crucial to measure circadian rhythms outside the lab. Wearable technologies have shown promise for this purpose, but their limited validation for research and clinical use remains an open challenge. We propose an approximation-based least-squares method that efficiently extracts daily physiological parameters, including circadian rhythms, from wearable data. Our method has been tested on simulated and real-world datasets, demonstrating accuracy and efficiency. It allows for identifying a reasonable harmonic model for circadian assessment, and surprisingly, a single-harmonic model outperforms a multiple-harmonic one. By integrating our method with a Kalman filtering algorithm, we estimate the clock state, making it a valuable tool for rigorous research and clinical application of wearables.

Speaker: Peter R. Kramer (Rensselaer Polytechnic Institute)

Title: Stochastic effects in molecular motor teams under detachment and reattachment

Abstract: We revisit two paradigms of cooperative action by kinesin molecular motors involving a coupling of the detachment and reattachment processes with the stochastic spatial dynamics. First, for two dissimilar types of kinesin transporting a common cargo, we provide approximate analytical characterizations for how incorporating slack in the tether model affects the cooperative dynamics. Secondly, we extend consideration of gliding assays to a situation where microtubules are crosslinked while being crowdsurfed by immobilized kinesin.

Speaker: Ryeongkyung (RK) Yoon (University of Houston)

Title: Identifying Interpretable Reaction Networks using Bayesian Inference.

Abstract: With the rapid increase of available observational data in Systems Biology, it becomes an important task to identify interpretable reaction networks governing biological systems. Because of the stochasticity from various noise sources and the complexity of biological processes, most existing deterministic frameworks are limited in extracting relevant information from massive data. In this work, we propose a framework based on the Stochastic Differential Equation (SDE) that explicitly accounts for variabilities and inherent correlations in time-series data. Using the Bayesian inference, we quantitatively compare two models in several different scenarios, such as data observed at the equilibrium state, measured at coarse frequency, and corrupted by both intrinsic and observational noise to illustrate outperformed accuracy and identifiability of the SDE-based model.

Speaker: Hyunjoong Kim (University of Pennsylvania)

Title: Social Foraging and Learning from Interactions

Abstract: Foraging is a fundamental behavior of nearly all animals, and learning environment from interactions is crucial for successful foraging. Unlike individual foragers, social foragers can learn environment from others’ experiences and impact the environment through mass interaction. In this talk, we will see how efficiency depends on group size and communication by introducing a Markov decision process analogous to a multi-armed multi-bandit problem. Interestingly, the stochastic model shows qualitatively different features from its deterministic limit when “buzzing” too much. We will also consider how various types of foragers (followers and dedicated searchers) can affect the efficiency under various food distributions. This talk is based on joint work with Joshua Plotkin and Yoichiro Mori.

Speaker: Jinsu Kim (POSTECH)

Title: Test three different Markov models for the Chlamydia developmental cycle

Abstract: Chlamydia is an intracellular bacterium that reproduces via an unusual developmental cycle such as late RB-EB conversion and heterogeneity of individual Chlamydia size. A key step is a conversion from a replicating form (RB) to an infectious form (EB), which occurs in a delayed and asynchronous manner. The regulatory mechanisms that control this developmental switch are unknown, but could potentially include extrinsic signals from the host cell or from other chlamydiae, or an intrinsic signal such as chlamydial cell size. In this presentation, we introduce three stochastic models, each based on a different regulatory mechanism. To test the models, we use the intrinsic noise of each model that can be estimated with statistical quantities measured experimentally. We found that all three models successfully reproduced the observed timing of RB-to-EB conversion and the growth curves of the developmental forms within an inclusion. However, only one model, based on the regulation of RB-to-EB conversion by RB size, was able to produce the positive correlation between the number of RBs and EBs and the monotonic time evolution of the coefficient of variation in the RB population.

Speaker: Gaoyang Fan (Altos Labs)

Title: Epigenetic Landscape of Cellular Reprogramming

Abstract: Waddington's landscape provides an intuitive visualization of how the interaction of genes and the environment shapes the cellular developmental trajectory toward different fates. Quantifying the epigenetic landscape has been done by construction of an energy function describing the cell state transition driven by the stochastic gene activity. Experimental evidence shows that over-expressing key regulators that maintain pluripotency leads to resetting of the chromatin state, resulting in a small subpopulation regaining stemness. To understand the path of cell reprogramming, we propose a dynamically changing quasi-potential landscape that integrates tools from dynamical systems and statistical mechanics. In particular, we construct a minimal epigenetic network and show how multiplicative noise arises from its chemical master equation may lead to different outcomes of cellular reprogramming.

Speaker: Seunggyu Lee (Korea University)

Title: Numerical aspect of quasi-steady state approximation for Michaelis-Menten reaction diffusion systems

Abstract: Quasi-steady state approximation (QSSA) are the well-known reduction method for chemical reaction models, but not the case with diffusion. In this talk, we discuss QSSA for Michaelis-Menten reaction-diffusion systems in a numerical aspect of view. It is mainly focused on the limited condition between standard QSSA and total QSSA, which was recently developed to accurate and efficient estimation of enzyme kinetic parameters for ODE systems, by comparing with the original full model.

Speaker: Tsvi Tlusty (IBS Center for Soft and Living Matter)

Title: Deciphering Proteins in Theory and Experiment

Abstract: Our starting point is the idea that specific regions in the protein evolve to become flexible viscoelastic elements facilitating conformational changes associated with function, especially allostery. Simple theories show how these regions can emerge through evolution and indicate that they are easily identified by amino acid rearrangement upon binding (i.e., shear motion). Surprisingly, AlphaFold can also identify such regions by computing the shear induced by a single or a few mutations. With these methods, we have tested the concept of shear and its functional relevance in various proteins. I will present recent results from an experimental study of the enzyme guanylate kinase linking shear, large-scale motions, and catalytic function. Looking at proteins as evolving viscoelastic machines is proposed as a predictive approach to understanding the basic principles of existing proteins and designing new ones.

Speaker: Yangjin Kim (Konkuk University)

Title: Role of NOTCH signaling in macrophage-mediated transport of intracellular molecules in OV therapy

Abstract: The impact of NOTCH signaling on immune therapy is understudied. Here we report that activation of NOTCH signaling promotes an MDSC enriched immune suppressive environment in brain tumors that limits the benefit from oncolytic immunotherapy. TCGA data analysis showed that brain tumor patients with high NOTCH signaling in their tumors had a higher macrophage infiltration, and also correlated with a poor immunological score. Using RNA sequencing, CHIP-PCR, and mathematical models we identified that infected tumor cells induced ADAMTS1 expression via RBP-j mediated canonical NOTCH signaling, which enhanced macrophage recruitment in tumors, inducing chemotactic movement toward the tumor mass. The Jag1 expressing macrophages created a feed forward loop in tumor microenvironment that amplified NOTCH signaling in tumor cells distant from sites of viral infection, leading to tumor growth and activating other signaling cascades within the tumor cells. Macrophages recruited to oHSV treated tumors induced CCL2 production via TLR activation. CCL2 induction was also observed in both murine models bearing brain tumors and in GBM patients treated with rQnestin34.5. Mathematical models and experimental evidences indicate the phenotypic switch towards an M2 phenotype that were immunosuppressive and induced tumor growth. Pharmacologic blockade of NOTCH signaling rescued this effect and promoted a CD8 dependent anti-tumor memory response that enhanced therapeutic efficacy of oHSV therapy. Using mathematical models (PDEs and hybrid approaches), we investigated the fundamental mechanism behind these and predicted the optimal dosing strategies against NOTCH signaling and tumor growth overall.

Speaker: Luonan Chen (Center for Excellence in Molecular Cell Science, Chinese Academy of Sciences)

Title: Dynamics-based data science for Biology

Abstract: I will present a new concept "dynamics-based data science" in AI applications of biology and medicine for studying dynamical processes and disease progressions, including dynamic network biomarkers (DNB) for early-warning signals of critical transitions, spatial-temporal information (STI) transformation for short-term time-series prediction, and partial cross-mapping (PCM) for causal inference among variables. These methods are all data-driven or model-free approaches but based on the theoretical frameworks of nonlinear dynamics. We show the principles and advantages of dynamics-based data-driven approaches as explicable, quantifiable, and generalizable. In particular, dynamics-based data science approaches exploit the essential features of dynamical systems in terms of data, e.g. strong fluctuations near a bifurcation point, low-dimensionality of a center manifold or an attractor, and phase-space reconstruction from a single variable by delay embedding theorem, and thus are able to provide different or additional information to the traditional approaches, i.e. statistics-based data science approaches. The dynamical-based data science approaches will further play an important role in the systematical research of various fields in biology and medicine as well as AI.

Speaker: Younghae Do (Kyungpook National University)

Title: Multistability of a nonsmooth atopic dermatitis system

Abstract: Atopic dermatitis (AD) has been known as the most common allergic inflammatory skin disease and its immunopathogenetic network is very complex. In this paper, to investigate the complexity of AD, we study a in silico AD model based on the mechanisms of AD disease pathogenesis, which is described by a nonsmooth system with three switches. We uncover a new oscillating behavior Os called Serious Oscillation, which can make a clear distinction between oscillating behaviors. It thus makes possible to classify AD attractors, which is very similar to AD clinical symptoms used in SCORAD index. A striking finding is that by investigating the existence and the stability of all found attractors on the parameter space for Barrier permeability κP and Immune responses αI, there exist many different types of bi- and multistability: four and five different types of bistability and multistability, respectively. By characterizing these different types of bi- and multistability, we finally conclude that the complexity of AD is caused by multistability detected in too much wide parameter ranges. In addition, we show a peculiar bifurcation phenomenon occurred in nonsmooth dynamical system. Our results suggest that the existence of multistability will make it possible to better understand the complexity of AD, which can be applicable in the development of new therapy strategies.

Speaker: Heyrim Cho (UC Riverside)

Title: Mathematical approaches to overcome limited temporal data in biomedical applications

Abstract: The aspect of limited temporal data is one of the many challenges when dealing with clinical data. The amount of data that can be practically collected in everyday patients during the therapy is very limited due to the financial cost and the patient’s burden. This motivates us to transfer the mathematical and computational models to meet the challenges in clinical data, before we use them to guide patient therapy via prediction. In this talk, I will discuss modeling approaches to tackle this problem. For instance, I will discuss a Bayesian information-theoretic approach to determine effective scanning protocols of cancer patients. We propose a modified mutual information function with a temporal penalty term to account for the loss of temporal data. The effectiveness of our framework is demonstrated in determining image scanning scheduling for radiotherapy patients.

Speaker: Jay Newby (University of Alberta)

Title: Dynamic self organization and microscale fluid properties of nucleoplasm

Abstract: The principal function of the nucleus is to facilitate storage, retrieval, and maintenance of the genetic information. A unique feature of nucleoplasm—the fluid of the nucleus—is that it contains chromatin (DNA) and RNA. In contrast to other important biological polymer hydrogels, such as mucus and extracellular matrix, the nucleic acid polymers have a sequence that encodes both genetic information and strongly influences spatial organization. How does crowding in a sequence specific hydrogel influence spatial organization of the dynamic molecular components responsible for nuclear function? We are becoming increasingly aware of the role of liquid-liquid phase separation (LLPS) in cellular processes in the nucleus and the cytoplasm. Complex molecular interactions over a wide range of timescales can cause large biopolymers (RNA, protein, etc) to phase separate from the surrounding nucleoplasm or cytoplasm into distinct biocondensates (spherical droplets in the simplest cases). I will discuss recent work modelling the role of nuclear biocondensates in neurodegenerative disease and several ongoing projects related to modelling and microscopy image analysis.

Speaker: Eunjung Kim (KIST)

Title: Resistance and Its Spatial Allocation Modulate Cancer Therapy Outcome

Abstract: Drug resistance is one of the leading causes of treatment failure in cancer therapy. We first explored the impact of preexisting resistance using a deterministic competition model and an agent-based model of the tumor cell population. Analyzing the ordinary differential equation (ODE) model that explains competition between sensitive and preexisting resistant cell populations, we showed that an effective dose window (EDW) might contain the tumor indefinitely. Simulation of our model shows that doses belonging to EDW can contain tumor progression indefinitely using either continuous or adaptive dose scheduling. Using optimal control theory, we showed that the lower bound of the EDW approximates the minimum effective dose (MED). However, the ODE model assumes homogeneous well-mixing of sensitive and resistant cells. In reality, cell configuration diverges from this homogeneity. To understand the impact of the resistant cell distribution, we developed an agent-based model. We simulated it for three tailor-made initial cell configurations: clumped, random, and uniform resistant cell distribution. The random and uniform cell distribution is initially close to the homogeneous mixing assumption of the ODE model and does not produce a significantly different result from the ODE model. The clumped distribution assumes a clump of resistant cells in the middle of the domain. Our analysis shows that under a continuous maximum tolerated dose (CT-MTD), tumors with clumped resistant cells end up with a longer progression time than tumors with random and uniform resistant cell distributions due to intra-species competition. When treated with adaptive therapy (AT), inter-species competition comes into effect in addition to intra-species competition and delays the progression further. So, the higher benefit of AT is associated with clumped initial cell configuration. Finally, we investigated the effect of acquired resistance on therapy outcomes with an agent-based model for spatial tumor growth. We considered three different types of acquired resistance: random genetic mutations during cell division, drug-induced reversible (plastic) phenotype changes, and drug-induced irreversible phenotype changes. These three resistance mechanisms led to different spatial distributions of resistant cells. To quantify the spatial distribution, we propose an extension of Ripley’s K-function, Sampled Ripley's K-function (SRKF), which calculates the non-randomness of the resistance distribution over the tumor domain. We predicted that the emergent spatial distribution of resistance could determine the time to progression under both AT and CT-MTD.

Speaker: Cheol-Min Ghim (UNIST)

Title: Taking Time to Catch Up: Beyond Steady-State Approximations in Chemical Kinetics

Abstract: The Michaelis-Menten (MM) rate law has enjoyed for over a century the status of the de facto standard of modeling enzymatic reactions. Despite its simple and intuitive interpretation for a wide range of applications in biochemistry, biophysics, cell biology, systems biology, and chemical engineering, the MM rate law and its modified form stand on the quasi-steady state assumption, which is not necessarily justified under active molecular concentration changes over time. Here, we relax this quasi-steady state requirement and propose the generalized MM rate law for the interactions of molecules with active concentration changes over time. Our approach for time-varying molecular concentrations, termed the effective time-delay scheme (ETS), is based on rigorously estimated time-delay effects in molecular complex formation. With particularly marked improvements in protein–protein and protein–DNA interaction modeling, the ETS provides an analytical framework to interpret and predict rich transient or rhythmic dynamics (such as autogenously-regulated cellular adaptation and circadian protein turnover), which goes beyond the quasi-steady state assumption.

Speaker: Paul Piho (Imperial College London)

Title: Quantifying the fitness effects of stochastic gene expression

Abstract: Cell-to-cell variation of gene expression in clonal populations directly contributes to natural selection when cells compete for growth. Stochastic modelling can help unravel and quantify the effects of noisy gene expression on fitness in cell populations. Chemical master equations are a well established framework for modelling intracellular gene regulation dynamics and recent advances use agent-based modelling of growing and dividing cells to extend these models to growing cell populations. However, these works often assume that the gene expression network of interest does not affect the cell division and growth. Here, we propose a new agent-based framework for gene expression systems coupled with cell division. Thus, noisy gene regulation in these systems directly contributes to natural selection. We compare the agent-based framework to effective master equation models where the effects of cell divisions are modelled implicitly through first-order effective decay or a dilution reaction. We use the developed theory to show that these effective models can be in qualitative disagreement with the agent-based description due to the effects of selection acting on the stochastic reaction network. Finally, we demonstrate parameter inference for the agent-based models using real-world data sets, unravelling and quantifying the effects of selection on the reaction network dynamics.

Speaker: Eui Min Jeong (IBS)

Title: Noise attenuation and ultrasensitivity in biological oscillators utilizing the multiple transcriptional repression mechanism

Abstract: In many biological systems, multiple repression mechanisms are used together to inhibit transcriptional activators in many systems. This raises the question of what advantages arise from utilizing multiple repression mechanisms. Here, by deriving Fano factors and equations describing the multiple repression mechanisms, we find that their combination can reduce noise in the transcription while generating an ultrasensitive transcription response and thus, strong oscillation. This rationalizes why multiple repression mechanisms are used in various biological oscillators.

Speaker: Hyun Kim (IBS)

Title: scLENS: Data-driven signal detection for scRNA-seq analysis

Abstract: Many dimensionality reduction tools have been developed to extract biologically meaningful signals from high-dimensional sparse and noisy scRNA-seq data. However, most of these tools leave the decision of selecting the optimal embedding dimension to user, introducing user bias into the analysis results. Furthermore, log normalization employed in these tools as a preprocessing step distorts signals, leading to false discovery. Here, we introduce scLENS, a dimensionality reduction tool to capture biological signals accurately by addressing these issues. In this study, we identified the cause of signal distortion due to conventional preprocessing method and minimized it using modified preprocessing method. Moreover, we employed the random matrix-based noise filtering and signal robustness test to provide a data-driven threshold differentiating signal from noise. By combining the signal correction after log normalization and the data-driven signal detection, scLENS outperformed ten widely used packages across 39 real and simulated datasets.

Speaker: Hyeonate Jo (IBS)

Title: Density physics-informed neural network infers an arbitrary density distribution for a non-Markovian system

Abstract: In this talk, we developed Density-PINN (Physics-Informed Neural Networks), a method capable of estimating the probability density function embedded within a differential equation. While conventional PINNs have focused on determining the solutions or parameters of differential equations that can explain observed data, we introduce a specialized approach for estimating the probability density function contained within the equation. Specifically, when dealing with a limited number of stochastic time series as observed data, and where only the average of the data satisfies the solution of the differential equation, we have constructed a mean-generating model using Variational Autoencoders. By applying our method to single-cell gene expression data from 16 promoters in response to antibiotic stress, we discovered that promoters with slower signaling initiation and transduction exhibit greater cell-to-cell heterogeneity in response intensity.

Speaker: Pan Li (IBS)

Title: Modeling the Circadian Control of Cardiac Function

Abstract: The circadian clock is a crucial biological mechanism that regulates diverse physiological processes in the body, including the cardiovascular system. In this talk, we will explore the intricate interactions between the circadian clock and cardiac function, and show how mathematical modeling can aid in comprehending these interactions. Specifically, we will focus on employing mathematical approaches to investigate the role of circadian rhythmicity in regulating heart rate and ventricular repolarization.

Speaker: Hyukpyo Hong (KAIST/IBS)

Title: Network translation allows for revealing long-term dynamics of stochastic reaction networks

Abstract: Long-term behaviors of biochemical systems are described by steady states in deterministic models and stationary distributions in stochastic models. Their analytic solutions can be obtained for limited cases, such as linear or finite-state systems. Interestingly, analytic solutions can be easily obtained when underlying networks have special topologies, called weak reversibility (WR) and zero deficiency (ZD). However, such desired topological conditions do not hold for the majority of cases. Thus, we propose a method of translating networks to have WR and ZD while preserving the original dynamics was proposed. Additionally, we prove necessary conditions for having WR and ZD after translation. Our method provides a valuable tool for analyzing and understanding the long-term behavior of biochemical systems, and we demonstrate its efficacy with several examples.