Poster Session - 2020

Hector Banos: “tRNA Neutral Networks in C. elegans”

Abstract: A key concept in evolutionary biology is fitness, but quantifying fitness is not a trivial task. To be able to identify deleterious or advantageous mutations, one needs to first understand the mutations that are neutral, that is, mutations that do not impact negatively nor positively fitness. Neutrality is tied to the principle of Genotype to Phenotype, where distinct genes associated with the same Phenotype are believed to be neutral. Here we investigate a classical approach to neutrality in RNA, particularly in tRNA in C elegans, via secondary structure prediction under the thermodynamic model.

Lianzhang Bao: “Vanishing-Spreading dichotomy in population invasion with chemotaxis”

Abstract: Predicting the evolution of expanding population is critical to controlling biological threats such as invasive species and cancer metastasis. In this work, we investigate the asymptotic dynamics in logistic type chemotaxis models in one space dimension with a free boundary. Such a model with a free boundary describes the spreading of a new or invasive species subject to the influence of some chemical substances in an environment with a free boundary representing the spreading front. Vanishing-Spreading dichotomy are first established in the chemoattraction-repulsion systems with a free boundary as well as with double free boundaries. The numerical simulations agree well with theoretical results and also indicate the dependence of the vanishing-spreading dichotomy on the initial total density, initial habitat, the moving speed and the chemotactic sensitivity coefficients.

Keisha Cook “Differentiating between single particle transport types”

Abstract: We are interested in the transport of lysosomes in human lung cells; some containing titanium dioxide nanoparticles and others not. Identifying and classifying the transport types involve understanding the velocity of the lysosomes, which is not reported in the trajectory data collected in the lab. Additionally, we may see switches in the transport types caused by a sudden change in velocity. For initial evaluation purposes, we choose frames rates that minimize the amount of switching behavior. For each trajectory, we estimate the velocity using a posterior distribution based on our statistical model. We fit the kernel density estimation to evaluate which trajectories exhibit velocities of immobile lysosomes versus freely diffusing lysosomes.

Daniel Cruz: “Agent-Based Modeling of Emergent Patterning and Canalization Within Pluripotent Stem Cell Colonies”

Abstract: Although various factors contribute to the transformation of pluripotent stem cell colonies into differentiated tissue, it is unclear which mechanisms take priority in context-specific situations. In this work we consider human induced pluripotent stem cells (hiPSCs), stem cells whose therapeutic potential arises from their ability to differentiate into all germ lineages. Using experimental data, we develop an agent-based model whose agents (cells) are each equipped with a Boolean network that governs how each cell interacts with its environment and with other cells. In this manner, we study both the dynamics of each Boolean network and the emergent features of the model as a whole. Our findings indicate that modeling and testing hiPSC colonies is a tractable approach to understanding how intercellular communication determines cell fate during early stages of differentiation.

Gemechis Degaga: “Generative Adversarial Neural Networks (GANs) for Targeted Protein Secondary Structures and IDPs”

Abstract: Molecular level understanding of the structure and function of proteins is a key part of modern biosciences and drug discovery research. Yet, vast majority known proteins contain intrinsically disordered regions also known as intrinsically disordered proteins (IDPs) for which stable structure and function is mostly not understood. Current computational modeling methods for protein design are slow and often require intensive human interventions. In addition, a major engineering challenge is to design new proteins in a controlled way for particularly targeted localized properties, folds, expressions, and functions. Here, we apply sequence based Deep Convolutional Generative Adversarial Networks (DC-GANs) to generate targeted protein secondary structures and IDPs, toward the understanding of phosphorylation mediated disorder-to-order transitions in fast de novo protein design. We encode our protein sequences with physicochemical and structural-statistical features associated to each residue. Our DC machine learning models with residue level local features showed robust secondary structure predictions and great versatility across the most common machine learning models. The effectiveness and accuracy of our models were confirmed by molecular dynamics and physical laboratory synthesis and evaluation using circular dichroism spectroscopy. We believe that this work could pave the way for the discovery of phospho-regulated disorder-to-order transition mechanism and impact the understanding of disorder related diseases and synthetic biology.

Parker Edwards & Nikola Milicevic: “Topological Data Analysis of Actin Networks”

Abstract: Networks of filaments assembled from the protein actin contribute significantly to cells' ability to move and change shape. These actin networks exhibit distinct local geometric structure. Some networks contain regions of straight and tightly packed fibers, for instance, as well as loops of varying sizes. Our data consist of high resolution live-cell microscopy images of cells' actin fibers from different classes of cells. Our methodology detects localized features using image segmentation and tools from topological data analysis: relative persistent homology, a novel approach in the field, and persistence landscapes. Using geometric summaries of each image as the feature vectors in machine learning algorithms, we produce several numerical scores for each image. Given hand-picked features and their scores for images coming from experts, we infer via regression what properties of the networks our scores are capturing.

Nolan English “A Machine Learning Recommendation for the Prediction of Functional Post-Translational Modifications”

Abstract: The modification of proteins after they are translated is an important process that can control the structure and function of the proteins on which they occur. Hundreds of different types of modification happen at some point during the lifetime of every protein in eukaryotic cells and play an essential role in cellular processes such as cell division, cell communication, gene regulation. Using current state-of-the-art detection tools, the rate at which post-translational modifications are detected now far surpasses the rate at which they can be investigated for functionality. Furthermore, not all modifications detected are functional, making it difficult to determine into which modifications one should invest experimental effort. Here, we describe a new computational tool – SAPH-ire TFx – capable of predicting functional modification sites from large-scale datasets, and consequently focus experimental effort towards only those modifications that are likely to be biologically significant. We show that the tool performs well across multiple datasets within which known functional modifications are scattered; and we show that the tool outperforms prior functional prioritization tools. Finally, we also provide a user-friendly web tool for experimentalists to investigate SAPH-ire TFx output for proteins and protein families of interest.

Margherita Ferrari: “R-loops role in RNA-templated DNA repair”

Abstract: In recent years, it has become evident that non-coding RNA can transfer information to DNA and impact gene expression. For instance, the Storici Lab showed that antisense transcript RNA can serve as a template for double-strand break (DSB) repair in yeast. The actual process of DSB repair is unclear, while experimental data indicate that repair must occur during transcription. We conjecture that DSB repair requires the formation of an R-loop, and we propose a testable mechanism where the antisense transcript RNA engages with the ssDNA of the R-loop formed by the sense transcript, and repairs the DNA. Moreover, we present mathematical approaches to model an R-loop, which in turn will help in analyzing RNA-DNA hybrids in the context of DSB repair.

MD Rafiul Islam: “An empirical comparison between integer and fractional order SEIR models of measles”

Abstract: We compare the performance of systems of ordinary and (Caputo) fractional differential equations depicting the susceptible-exposed-infected-recovered (SEIR) models of diseases. We examine the validity of the two approaches against empirical courses of epidemics, we fit both of them to case counts of three measles epidemics that occurred during the pre-vaccination era in three different locations. We found that while the fractional order differential equations (FDE) SEIR model gives a slightly better fit to some of the data, the ordinary differential equations (ODE) SEIR model performs better overall.

Benjamin Jones: “Comparing the Effectiveness of ESES and MSMS for Solving the Poisson-Boltzmann Equation”

Abstract: The solvation energy of a molecule is the energy absorbed or released when that molecule is dissolved in a solvent. Computational biologists use the solvation energy of macromolecules, such as proteins, to determine their stability for protein folding and their hydrophobicity for drug delivery.  The solvation energy of a protein can be found by solving the nonlinear Poisson-Boltzmann equation (PBE). The PBE is computationally intensive to solve for large proteins. Therefore, efficient, accurate, and stable algorithms are desired. The computation requires knowledge of a molecular surface. In this work, the Solvent Excluded Surface (SES) is used in the Poisson-Boltzmann model. The generation of this surface may lead to instabilities in the computation, particularly when using values that have a jump condition at the protein interface.
A standard SES generation tool is MSMS, which is known to cause some numerical instabilities. The Eulerian Solvent Excluded Surface (ESES) tool was recently created and may address some of these instabilities. In this work, the impact of the MSMS and ESES surfaces on numerical methods for solving the PBE are compared.

Jinsu Kim: “First passage time of a stochastic epigenetic model for chromatin accessibility characterizes rules by which transcription factor dynamics alter the epigenome”

Abstract: First passage time of a stochastic epigenetic model for chromatin accessibility characterizes rules by which transcription factor dynamics alter the epigenome. Jinsu Kim and Katherine Sheu. In macrophages, upon stimulation with pathogen or cytokine ligands, signal dependent transcription factors (SDTF) can demonstrate various dynamic temporal profiles, including their amplitude, duration, and oscillations, but it remains unclear how information contained in signaling dynamics is decoded by chromatin to mediate chromatin remodeling and impact cellular epigenetic states. In this project, we study stochastic models of epigenome dynamics, combined with genomic sequencing analysis, that allow predictive understanding of how nucleosomes at specific genomic locations respond to different patterns of SDTF signaling to produce alterations to chromatin states in health and disease conditions.

Bo Lin: “Optimizing coordinate choice for locomotion systems with shape spaces in non-simply-connected 2-manifolds”

Abstract: Geometric mechanics, a gait design framework, is widely used to study the limbless locomotion. In the geometric mechanics framework, the kinematic motion is prescribed by a trajectory in a parameterization of a shape space and such prescription determines the displacement in a position space. Often, the parallel parking effect (lateral displacement resulted from the combined effect of the forward and rotational velocities) can lead to the prediction-experiment discrepancy. It has been shown that such parallel effect can be minimized on systems in simply connected spaces. Recent work in geometric mechanics extended its scope to more complicated systems, such as the multi-legged systems where their shape spaces is contained in tori. While being effective to design gaits, the geometric mechanics on non-simply-connected 2-space often fail to quantitatively predict the displacement, partially because of the parallel parking effect. In this work, we developed a tool to minimize the parallel parking effect on the non-simply-connected 2-manifold. To address the non-simply-connectedness, we adapted the forms of the boundary conditions and numerically obtained the optimal coordinate using the finite-element method. We applied our methods to the centipede locomotion system and observed the quantitative agreement between geometric mechanics prediction and the experiment result.

Rohan Mehta: “Modeling anti-vaccine sentiment as a cultural pathogen”

Abstract: Culturally transmitted traits that have deleterious effects on biological, health-related traits in individuals can be regarded as cultural pathogens. A cultural pathogen can produce coevolutionary dynamics with its associated health-related traits, so that understanding the dynamics of a health-related trait benefits from consideration of the dynamics of the associated cultural pathogen. Here, we treat anti-vaccine sentiment as a cultural pathogen, modeling its “infection” dynamics together with the infection dynamics of the associated infectious disease against which the vaccine protects. In a coupled Susceptible-Infected- Resistant (SIR) compartmental model, consisting of an SIR model for the anti-vaccine sentiment and an interacting SIR model for the infectious disease, we explore the effect of anti-vaccine sentiment on infectious disease dynamics. We find that disease endemism is strongly contingent on the presence of the anti-vaccine sentiment. Furthermore, the dynamics of anti-vaccine sentiment can create situations in which the disease nears extinction but returns suddenly after a long period of dormancy. We study the effect of assortative sentiment-based interactions on the dynamics of sentiment and disease, identifying a tradeoff whereby assortative meeting aids the spread of a disease but hinders the spread of anti-vaccine sentiment. Our results can contribute to the identification of strategies for reducing the impact of a cultural pathogen on disease infection, illuminating the value of a cultural evolutionary modeling approach in the analysis of phenomena that affect disease dynamics.

Hamid Mofidi: “Effects of diffusion coefficients in ionic channels”

Abstract: We study the dependence of reversal potentials and zero-current fluxes on diffusion coefficients for ionic flows through membrane channels. We consider two mobile ion species, one positively charged (cation) and one negatively charged (anion). Numerical observations are obtained from analytical results established using geometric singular perturbation analysis of classical Poisson-Nernst-Planck models. We present some results with numerical observations for biological relevant situations. We show the numerical investigations on profiles of the electrochemical potentials, ion concentrations, and electrical potential across ion channels for the zero-current case. The dependencies of current and fluxes on voltages and permanent charges are also investigated.

Nida Obatake: “Hopf bifurcations in a model of ERK regulation”

Abstract: In this poster we present an algorithm to construct a positive point that solves a system of polynomial inequalities. The procedure uses ideas from polyhedral geometry, specifically normal fans of the Newton polytopes of the polynomials. We give an application of this algorithm to a problem from chemical reaction network theory, an area of mathematics that analyzes the behaviors of chemical processes. A major problem in this area is the stability of equilibria of dynamical systems arising from these networks. Here, we focus on a biological signaling network called the ERK network, a model for dual-site phosphorylation and desphorylation of extracellular signal-regulated kinase. The ERK network is known to be bistable and to exhibit oscillations (Rubinstein, Mattingly, Berezhkovskii and Shvarstman, 2016), but a limiting network of the ERK network- the fully processive dual-site network- is known to have a unique, stable steady state (Conradi and Shiu, 2015). We investigate the emergence of oscillations and instability by analyzing certain subnetworks of ERK. A precursor to oscillations is the existence of a Hopf bifurcation, which are characterized by sign conditions on Hurwitz-matrix determinants (Yang, 2002). Thus, finding a Hopf bifurcation amounts to finding a positive solution to a system of polynomial inequalities.  We present a solution using our algorithm.

Seth O’Conner: “Mitochondrial fostering: the mitochondrial genome may play a role in plant orphan gene evolution”

Abstract: Plant mitochondrial genomes exhibit odd evolutionary patterns. They have a high rearrangement but low mutation rate, and a large size. Based on massive mitochondrial DNA transfers to the nucleus as well as the mitochondrial unique evolutionary traits, we propose a “Mitochondrial Fostering” theory where the organelle genome plays an integral role in the arrival and development of orphan genes (genes with no homologues in other lineages). Two approaches were used to test this theory: 1) bioinformatic analysis of nuclear mitochondrial DNA (Numts: mitochondrial originating DNA that has migrated to the nucleus) at the genome level, and 2) bioinformatic analysis of particular orphan sequences present in both the mitochondrial genome and the nuclear genome of Arabidopsis thaliana. One study example is given about one orphan sequence that codes for two unique orphan genes: one in the mitochondrial genome and another one in the nuclear genome. DNA alignments show regions of this A. thaliana orphan sequence exist scattered throughout other land plant mitochondrial genomes. This is consistent with the high recombination rates of mitochondrial genomes in land plants. This may also enable the creation of novel coding sequences within the orphan loci, which can then be transferred to the nuclear genome and become exposed to new evolutionary pressures. Our study also reveals a high correlation between mitochondrial DNA rate transferred to the nuclear genome and number of orphan genes in land plants. All the data suggests the mitochondrial genome may play a role in nuclear orphan gene evolution in land plants.

Tracey Oellerich: “Adaptability Conditions in Biological Networks”

Abstract: In this work we extend adaptability conditions for biological networks to include singular systems with non-hyperbolic equilibria. The proposed theoretical extension is compatible with the notions of homeostasis and robust perfect adaptation (RPA) and clarifies the relationship between the two. The new condition is derived using the notion of Moore-Penrose pseudoinverse and is implemented using a numerically efficient algorithm. The proposed approach is tested on several synthetic systems that are shown to exhibit homeostatic behavior yet lie outside of the scope of earlier work.

Alexander Ruys de Perez: “A New Criterion for Nonconvexity in Neural Codes”

Abstract: A place cell is a neuron corresponding to a subset of Euclidean space known as a place field, that will fire if and only if the individual possessing the place cell is within that place field. The firing patterns of a collection of n place cells can be represented by a neural code C on n neurons, which is a subset of the power set of {1,2,...,n}. Determining whether C is convex, meaning that there is an arrangement of convex open place fields for which C is the code, remains an open problem.

Previous work has introduced the notion of a local obstruction, a phenomenon that guarantees a code to be nonconvex. However, the absence of local obstructions does not guarantee convexity; there are both convex and nonconvex codes that do not contain local obstructions. Here, we introduce a new criterion for nonconvexity, which we call a wheel, which can exist in codes that possess no local obstructions. We define the conditions that certify the existence of a wheel in a code, explain why it proscribes convexity, and show how it contributes toward a classification of convex and non-convex codes on six neurons.

Tracy Stepien: “Spreading Mechanics and Differentiation of Astrocytes During Retinal Development”

Abstract: In embryonic development, formation of the retinal vasculature is critically dependent on prior establishment of a mesh of astrocytes. Astrocytes emerge from the optic nerve head and then migrate over the retinal surface in a radially symmetric manner and mature through differentiation.  We develop a PDE model describing the migration and differentiation of astrocytes, and numerical simulations are compared to experimental data to assist in elucidating the mechanisms responsible for the distribution of astrocytes.

Ling Wang: “Mutational variation and RNA folding in natural populations”

Abstract: Single stranded RNA forms complex secondary structure via intricate base-pairing interactions. These secondary structures impart catalytic, ligand binding, and scaffolding features to a broad array of RNAs, such as miRNA, forming an integral node of biological regulation. Among their many functions, RNA structural constructions modulate mRNA translation, alternative splicing, and miRNA targeting. Those functions of RNA secondary structure might contribute to the trait expression and model the evolution of proteins.

Siwen Wang: A regularization approach for solving Poisson’s equation with singular charge sources and diffuse interfaces

Abstract: In this poster, a simple Poisson’s equation involving inhomogeneous media with a diffuse interface is studied. In particular, we will assume constant dielectric values inside each dielectric medium, while the dielectric function varies smoothly from one medium to another, through a narrow transition band. For Poisson’s equation with singular charges and diffuse interfaces, a semi analytical method have already been proposed. The singular charges are treated analytically in this approach with diffuse interfaces. Nevertheless, this method is limited to simple geometries. Meanwhile, in regularization methods, a Poisson equation with the same singular sources, the singular component can be analytically solved as Coulomb potentials or Green’s functions. the other potential components can be accurately solved by finite difference or finite element methods. However, all existing regularization methods are designed for piecewise constant dielectric functions with sharp interfaces. It is unclear if regularization formulation could be established for diffuse interfaces. Thus, this poster presents the first regularization method in the literature that is able to handle diffuse interfaces. Besides a decomposition of potential function, the success of the new method lies in a decomposition of the inhomogeneous dielectric function. The singular charge sources containing in a complex domain can then be analytically treated. The details of the proposed regularization formulation and numerical validation of a simple example will be discussed.

Keren Zhang: “Computational pipeline for quantifying synapses in C. elegans’s connectome”

Abstract: The connectome of C. elegans is the first comprehensive map of synaptic connections for an organism. Studying C.elegans's connectome is essential in understanding how neurons are connected structurally and how the connections are subject to change such as development. However, traditional methods of studying synapse mainly rely on manual scoring on a few samples, and thus lacking quantification and variability. With microfluidic chips and fluorescent microscopy, we acquired massive amount of data to quantify the synapses and include individual variation. we propose a machine learning-based computational pipeline to extract the synapses's quantitative phenotypes from 3D confocal fluorescent microscopy images, we propose a computational pipeline. Due to the low signal to noise ratio and data imbalance, we first use a convolutional neural network (UNet) to segment out region of interest (ROI), which is neurite of presynaptic and postsynaptic neurons that co-localize with synapses. Then we further segment out synapses and subsequently extract quantitative phenotypes. Our computational pipeline is end-to-end and only requires minimum amount hand curation. We demonstrate that this pipeline can capture synaptic difference between different gender and throughout development


A NSF-Simons MathBioSys Research Center