Statistical challenges and opportunities in infectious disease modeling

Date Wednesday January 10, 2018 at 1:00 PM
Location 13-105 Center for the Health Sciences
Speaker Vladimir Minin, Ph.D., Professor of Statistics, University of California, Irvine
Sponsoring Dept UCLA Biomathematics
Abstract Stochastic epidemic models describe how infectious diseases spread through a population of interest. These models are constructed by first assigning individuals to compartments (e.g., susceptible, infectious, and recovered) and then defining a stochastic process that governs the evolution of sizes of these compartments through time. I will discuss multiple strategies for fitting these models to data, which turns out to be a challenging task. The main difficulty is that even the most vigilant infectious disease surveillance programs offer only noisy snapshots of the number of infected individuals in the population. I will discuss Bayesian data augmentation strategies that make statistical inference with stochastic epidemic models computationally tractable. Some of these strategies can even handle more exotic data types, such as infectious disease agent genetic sequences collected during outbreak monitoring. I will present results of fitting stochastic epidemic models to data from outbreaks of influenza and Ebola viruses.
Flyer 20180110_Vlad_Minin_flyer_(2).pdf

Mathematical Models for Mechanisms Driving Asymmetric Cell Division

Date Thursday November 30, 2017 at 4:00 PM
Location 13-105 CHS Center for the Health Sciences
Speaker Blerta Shtylla, Ph.D., Assistant Professor, Department of Mathematics Pomona College
Sponsoring Dept UCLA Biomathematics
Abstract In early development, embryo cells can opt to divide asymmetrically and produce daughters that have distinct fates. At the cell scale, asymmetric division is achieved with the help of exquisitely fine-tuned mechanistic processes that control the location of the cell division plane. In this talk we will introduce two mechanisms for how a dividing cell might achieve asymmetric division. First, we start with the simple single cell bacteria, Caulobacter Crescentus and show how a single cell employs spatio-temporal protein localization patterns to estimate the location of the division site. Using advection-reaction-diffusion PDE models we show how modulation of ATP-ase reaction rates can affect the spatial location of several proteins and transition the cell from a symmetric to an asymmetric division program. We also show how Brownian dynamics simulations can be used to verify some of the conclusions of the PDE models. A second example of how asymmetric division is achieved will be illustrated in the early stages of C. elegans embryo division. In this case, the cell employs more complicated biopolymer networks to achieve proper asymmetric division plane placement. We use stochastic models and experimental data to show how division plane placement can be controlled robustly. In both cases, model results will be compared with data and general mechanisms for division plane placement will be discussed.
Flyer 20171130_Blerta_Shtylla_flyer.pdf

How B Lymphocytes Evolve in Programmed and Adaptive Antigenic Environments

Date Thursday November 16, 2017 at 4:00 PM
Location 13-105 CHS (Center for the Health Sciences)
Speaker Shenshen Wang, Ph.D., Department of Physics & Astronomy, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract Clusters of proliferating B cells emerge and compete fiercely for survival in secondary lymphoid tissues upon exposure to antigen. Rapidly mutating pathogens (notably HIV) challenge natural immunity and vaccine design alike; a protective vaccine requires eliciting antibodies capable of neutralizing most antigen variants, thus named broadly neutralizing antibodies (bnAbs). However, in natural settings, bnAbs develop rarely and only after a long and complex B cell maturation process. A key gap in understanding remains between bnAb precursor activation and maturation completion. Using statistical physics models and multi-scale numerical realizations, we study dynamics of B cell affinity maturation in response to predefined discrete antigens and continuously co-evolving ones, respectively, considering complexity of antigen and fluctuations in B cell populations. We identify what differ in the means by which broad neutralization can be achieved in these settings, which in turn reveal intrinsic and external factors that shape adaptation pathways and dictate evolution outcomes.
Flyer 20171116_Shenshen_Wang_flyer.pdf

Evolutionary Dynamics of Cancer and Its Response to Treatment

Date Thursday November 09, 2017 at 4:00 PM
Location 13-105 CHS (Center for the Health Sciences)
Speaker Ivana Bozic, Ph.D., Dept. of Applied Mathematics, Univ. of Washington
Sponsoring Dept UCLA Biomathematics
Abstract Cancer is the result of a stochastic evolutionary process characterized by the accumulation of mutations that are responsible for tumor growth, immune escape, and drug resistance, as well as mutations with no effect on the phenotype. Stochastic modeling can be used to describe the dynamics of tumor cell populations and to obtain insights into the hidden evolutionary processes leading to cancer. I will present recent approaches that use branching process models of cancer evolution to quantify intra-tumor heterogeneity and the development of drug resistance, and their implications for interpretation of cancer sequencing data and the design of optimal treatment strategies.
Flyer 2017119_Ivana_Bozic_flyer.pdf

Stochastic Compartmental Models of Infectiuos Disease Without Tiresome Simulation or Gross Approximation

Date Thursday November 02, 2017 at 4:00 PM
Location 13-105 CHS (Center for the Health Sciences)
Speaker Marc Suchard, M.D., Ph.D., Professor, Departments of Human Genetics and Biomathematics
Sponsoring Dept UCLA Biomathematics
Abstract Stochastic compartmental models are important tools for understanding the course of infectious diseases epidemics in populations and in prospective evaluation of intervention policies. However, calculating the likelihood for discretely observed data from even simple models – such as the ubiquitous susceptible-infectious-removed (SIR) model – has been considered computationally intractable, since its formulation almost a century ago. Recently researchers have proposed methods to circumvent this limitation through data augmentation or approximation, but these approaches often suffer from high computational cost or loss of accuracy. We develop the mathematical foundation and an efficient algorithm to compute the likelihood for discretely observed data from a broad class of stochastic compartmental models. The computational complexity scales polynomially with the changes in population sizes between observations. We achieve this through a novel re-parameterization of the stochastic compartmental model into a multivariate coupled birth process and identify a convenient recursion to express the Laplace transform of the finite-time transition probabilities. We also give expressions for the derivatives of the transition probabilities using the same technique, making possible inference via Hamiltonian Monte Carlo (HMC). We use the 17th century plague in Eyam, a classic example of the SIR model, to compare our recursion method to sequential Monte Carlo, analyze using HMC, and assess the model assumptions. We also apply our direct likelihood evaluation to perform Bayesian inference for the 2014-2015 Ebola outbreak in Guinea under a hierarchical and time-inhomogeneous extension of the SIR model. The results suggest that the epidemic infectious rates have decreased since October 2014 in the Southeast region of Guinea, while rates remain the same in other regions, facilitating understanding of the outbreak and the effectiveness of Ebola control interventions
Flyer 20171102_Marc_Suchard_flyer.pdf

Multi-omics Integration and Network Modeling to Understand Complex Metabolic and Brain Disorders

Date Thursday October 19, 2017 at 4:00 PM
Location 13-105 CHS (Center for the Health Sciences)
Speaker Xia Yang, Ph.D., Associate Professor Dept. of Integrative Biology & Physiology UCLA
Sponsoring Dept UCLA Biomathematics
Abstract ABSTRACT: Complex human diseases such as cardiovascular disease, diabetes, and neurodegenerative disorders involve multi-scale alterations and re-organization at molecular, cellular, and tissue levels. Technological advances in the past decade are enabling an outpour of multi-omics big data to assist with our understanding of complex human diseases, but also pose challenges to multi-scale data integration. In this talk, I will introduce i) systems genomics approaches and tools to accommodate diverse omics data types from different species for network modeling of pathophysiological systems, and ii) applications of these methods in metabolic and brain disorders. These efforts have led to the elucidation of cell-specific and tissue-specific networks and critical regulators either specific to individual diseases or shared between diseases. The systems and network level insights can help guide the development of novel network-based medicine for common complex human diseases.
Flyer 20171019_Xia_Yang_flyer.pdf

oupling a Continuum Model and Live Imaging to Infer Tissue Spreading Mechanics: A Bayesian Approach

Date Thursday October 12, 2017 at 4:00 PM
Location 13-105 CHS Center for the Health Sciences
Speaker Tracy Stepien, Ph.D., Postdoctoral Research Associate Department of Mathematics University of Arizona
Sponsoring Dept UCLA Biomathematics
Abstract Collective cell migration is a major contributor to embryonic development, wound healing, and the progression of many diseases, and it has been successfully simulated using a range of modeling formalisms. If different types of collective cell migration are driven by a set of shared behaviors, a mathematical model of collective migration should successfully model different tissues with only a change in parameters. Here, we extend a two-dimensional Eulerian continuum mechanical model of a spreading tissue that was previously developed for cultured epithelial cell sheet migration in combination with quantitative image analysis to describe collective migration of embryonic tissue explants excised from the animal cap region of gastrulating Xenopus laevis embryos. This model assumes that the main parameters that influence collective migration are the forces on the free edge, tissue stiffness, and the strength of cell-ECM adhesions. We apply an automated methodology using approximate Bayesian computation to integrate kinematic data with an appropriately constrained computational model to predict physical properties of collectively spreading cell sheets. Our results suggest both the force of lamellipodia and cell-ECM adhesion vary with the initial area of the tissue explants; in particular, high lamellipodial force together with weaker cell adhesion enable fast spreading in explants with a larger initial area. Such predictions can be used to guide further experiments to better understand how collective migration is regulated during development and dysregulated during the metastasis of cancer.
Flyer 20171012_Tracy_Stepien_flyer.pdf

Stochastic Models for Gene Expression: Volume Growth, Cell Division, Gene Replication and Interactions Between Protein Productions

Date Thursday October 05, 2017 at 4:00 PM
Location 13-105 CHS (Center for the Health Sciences)
Speaker Renaud Dessalles, Ph.D. , Visiting Asst Proj Scientist Department of Biomathematics UCLA
Sponsoring Dept UCLA Biomathematics
Abstract Protein production is the fundamental process by which the genetic information of a cell is synthesized into functional products, the proteins. It is a highly stochastic process, in particular for the bacteria, as it is the realization of a very large number of elementary random events of different nature. Classical models of protein production (like those of Rigney and Schieve (1977) and Berg (1978) represent transcription and translation mechanisms to determine their relative impact on the protein variability. Even if these models are commonly used in the literature, they do not represent many aspects yet possibly impacting the protein variability: for instance, the volume growth, the cell division, the gene replication and the sharing of common resources such as RNA-polymerases and ribosomes in the protein synthesis are not represented. We propose here a series of models that successively integrates all these elements; mathematical analysis and simulations of these models will allow us to determine the variability induced by these different features and to compare them to experimental measures.
Flyer 20171005_Renaud_Dessalles_flyer.pdf

Functional Regression Models for Gene-based Association Studies of Complex Traits

Date Thursday May 04, 2017 at 4:00 PM
Location 23-105 CHS (Center for the Health Sciences )
Speaker Ruzong Fan, Ph.D., Professor, Department of Biostatistics, Bioinformatics, and Biomathematics, Georgetown University
Sponsoring Dept UCLA Biomathematics
Abstract By using functional data analysis techniques, fixed effect functional regression models are developed to test associations between complex traits and genetic variants, which can be rare variants, common variants, or a combination of the two, adjusting for covariates. We treat multiple genetic variants of an individual in a human population as a realization of an underlying stochastic process. The genome of an individual is viewed as a stochastic function which contains both genetic position and linkage disequilibrium (LD) information of the genetic markers. To overcome the curse of high dimensions of modern genetic data, functional regression models are developed to reduce the dimensionality. In the talk, I will show how to build test statistics for functional regression models to test association between quantitative/dichotomous/survival traits and genetic variants. Results of extensive simulation analysis and real data analysis will be shown to demonstrate the performance of the proposed models and tests. A comparison with existing popular procedure of sequence kernel association test (SKAT) and its optimal unified test (SKAT-O) will be made to facilitate an understanding of the proposed methods, and to answer whether fixed or mixed models should be used in association analysis of complex disorders.
Flyer Ruzong_Fan_20170504.pdf

Science for the greater good: how a Math professor saved the Italian coastline from Big Oil

Date Tuesday April 18, 2017 at 4:00 PM
Location 13-105 CHS Center for the Health Sciences
Speaker Maria D’Orsogna, Ph.D., Professor, Mathematics Department, CSU Northridge; Associate Adjunct Professor, Biomathematics Department, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract Trained as a physicist, Dr. Maria D’Orsogna is a hands-on activist fighting to protect the Italian countryside and seas from Big Oil. In 2007 she learnt of proposed oil activities in her home region of Abruzzo, Italy where century-old wineries were to be uprooted to build wells, refineries and pipelines, turning scenic Abruzzo into an oil district. Although based in California - 6,000 miles away - Dr. D’Orsogna took it upon herself to raise awareness and educate the public at large. She blended her scientific training, her experience as a professor, and her strong desire for social justice into an environmental movement that rapidly spread from Abruzzo across the country. Over the years, she traveled from town to town in Italy, educating citizens about environmental and health effects, debating Big Oil, exposing political corruption, engaging the Catholic Church, putting pressure on decision makers to act for the common good. While in California she used social networks and blogging to expose wrongdoings of the oil and gas industry, coordinate letter writings, keep raising awareness and spur action. Thanks to public uproar, spearheaded by Maria’s unwavering efforts, Abruzzo banned oil drilling and for the first time ever, in 2016, the Italian parliament imposed a no-drill zone of 12 miles encompassing all of Italy’s 5,000 mile coastline. Overall she helped stop 40 oil leases, earning the nickname “Erin Brockovich of Italy”. Maria’s story is a testament of how, by engaging with the community, scientists and educators can truly make a difference.
Flyer Maria_DOrsogna_20170418_revised2.pdf

Thermodynamics of Error Correction

Date Friday January 13, 2017 at 4:15 PM
Location A2-342 MDCC, Moss Auditorium, Marion Davies Children Clinic
Speaker Simone Pigolotti, Ph.D., Guest Scientist, Max-Planck Institute for Complex Systems Dresden, Germany
Sponsoring Dept UCLA Biomathematics
Abstract ABSTRACT: Biological systems are able to replicate information with outstanding accuracy. In biochemical reactions, such as DNA duplication, different monomers can be distinguished because of their binding energies or via non-equilibrium kinetic mechanisms. I will show how, in simple copying reactions, these two discrimination modes are mutually exclusive and lead to opposite tradeoffs between error, dissipation and reaction velocity. In multi-step reactions, such as in kinetic proofreading, these different modes can be combined to improve overall accuracy. I will conclude by discussing how the second law of thermodynamics can be used to directly relate copying accuracy with thermodynamic observables.
Flyer Simone_Pigolotti_seminar_20170113.pdf

New Rules of Inheritance in Ancient Animals

Date Thursday December 01, 2016 at 4:00 PM
Location A2-342 MDCC, Moss Auditorium, Marion Davies Children Clinic
Speaker Bo Wang, Ph.D., Assistant Professor, Department of Bioengineering, Stanford University
Sponsoring Dept UCLA Biomathematics
Abstract We humans reproduce sexually, with inheritance accomplished through a combination of DNA from our parents. However, not all lifeforms operate under this mode of inheritance. Many basal animals reproduce asexually; their progeny directly inherit unadulterated parental DNA. Mutations that accumulate in the parental somatic tissue can be passed down to its descendants. Because they are unable to eliminate undesirable mutations through genetic recombination like sexually reproducing animals can, asexual lifeforms are often considered an evolutionary dead end. Similar mechanisms also underlie important health-related problems, such as the emergence of drug resistance in tumors, the pathological evolution of the gut microbiome, and even the directed evolution of microbes for metabolic engineering. With the important observation that a small group of asexual animals have evolved cryptic strategies to outrun this dead end, we study an extraordinary system of such, the planarian, small invertebrates renowned for their incredible regenerative capacity. In this talk, I will introduce the biology underlying this poorly understood mode of inheritance, and specify areas in which mathematics is needed to make progress.
Flyer Bo_Wang_20161201.pdf

Multistationarity in Biochemical Reaction Networks

Date Friday November 18, 2016 at 3:00 PM
Location 13-105 CHS, Center for the Health Sciences
Speaker Badal Joshi, Ph.D., Assistant Professor, Department of Mathematics, California State University, San Marcos
Sponsoring Dept UCLA Biomathematics
Abstract The dynamics of a biochemical reaction network are frequently modeled using ordinary differential equations containing numerous unknown parameters (reaction rate constants). The values of these rate constants are difficult to determine experimentally. It is imperative, therefore, to develop theory that can determine the range of qualitative behaviors that a network can possess without requiring detailed knowledge of the parameters. In this talk, I will focus on the existence (or absence) of multiple positive steady states for a reaction network, also known as multistationarity. Existence of multiple positive steady states provides the underpinnings for switching in biochemical reaction networks, while uniqueness of a positive stable steady state is necessary for robust system output. In recent years, there have been multiple advances on the fronts of both ruling out multistationarity and establishing multistationarity in reaction networks. The theoretical results fall under three major areas: (a) deficiency theory, (b) injectivity (Jacobian criteria), (c) lifting/network embedding. I will describe the main ideas in the three areas along with some of the more recent results.
Flyer Badal_Joshi_20161118.pdf

Stochastic Modeling and Inference with Multi-type Branching Processes

Date Thursday November 17, 2016 at 4:00 PM
Location A2-342 MDCC, Moss Auditorium
Speaker Jason Xu, Ph.D., Postdoctoral Fellow, Department of Biomathematics, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract Markov branching processes are a class of continuous-time Markov chains (CTMCs) with ubiquitous scientific modeling applications. Multi-type processes are necessary to model phenomena such as competition, predation, or infection, but often feature large or uncountable state spaces, rendering standard CTMC techniques impractical. We present recent methodology that enables likelihood computations and EM algorithms in these settings. We examine the performance of these techniques applied to data from molecular epidemiology and hematopoiesis studies, and briefly explore alternatives when such methods are limited, including moment-based estimators and compressed sensing techniques that scale to large systems and datasets.
Flyer Jason_Xu_20161117.pdf

Better Living Through Control: With Applications to Neural and Cardiac Systems

Date Thursday November 10, 2016 at 4:00 PM
Location A2-342 MDCC, Moss Auditorium, Marion Davies Children Clinic
Speaker Jeff Moehlis, Ph.D., Professor, Department of Mechanical Engineering, University of California, Santa Barbara
Sponsoring Dept UCLA Biomathematics
Abstract Some brain disorders are hypothesized to have a dynamical origin; in particular, it has been been hypothesized that some symptoms of Parkinson’s disease are due to pathologically synchronized neural activity in the motor control region of the brain. We have developed a procedure for determining an optimal electrical deep brain stimulus which desynchronizes the activity of a group of neurons by maximizing the Lyapunov exponent associated with their phase dynamics, work that could lead to an improved method for treating Parkinson’s disease. The use of related control methods for treating other medical disorders, including cardiac arrhythmias such as alternans, will also be discussed.
Flyer Jeff_Moehlis_20161110.pdf

Cellular Variability and Information Flow in Signal Transduction Networks

Date Thursday November 03, 2016 at 4:00 PM
Location A2-342 MDCC, Moss Auditorium, Marion Davies Children Clinic
Speaker Roy Wollman, Ph.D., Assoc. Prof., Dept Integrative Biology & Physiology. Dept Chem. & Biochem., Inst. for Quantitative & Computational Biology, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract Signaling networks act as sensors, or measurement devices, that provide information on the extracellular environment to allow cells to respond to environmental changes appropriately. Experimental single cell measurements of signaling responses indicated high level of response variability raising the possibility that cellular responses are limited in their biochemical accuracy. I will discuss our efforts to examine the question of the accuracy of cellular signal transduction networks. I will show how cells utilize temporal signal modulation--that is, dynamics--to reduce noise-induced information loss and increase the accuracy of cellular response. In the context of wound response signaling, I will discuss how cells communicate with each other optimally to allow for “local averaging” that increases information about their position relative to the wound. Finally, I will show that cellular population is composed of mixtures of different cellular states, and that the existence of multiple cellular states explains some of the observed cell to cell variability. Through the use of mixture of multiple classes of multivariate cellular responses combined with paracrine information sharing among cells, a cellular population can increase its response appropriately to environmental changes.
Flyer Roy_Wollman_20161103.pdf

Interactions within Multispecies Microbial Networks

Date Thursday October 27, 2016 at 4:00 PM
Location A2-342 MDCC, Moss Auditorium Marion Davies Children Clinic
Speaker James Boedicker, Ph.D., Assistant Professor, Physics and Astronomy and Biological Sciences, University of Southern California
Sponsoring Dept UCLA Biomathematics
Abstract Microbial ecosystems in nature are diverse and heterogeneously distributed in space and time. Interactions between the microorganisms within these communities help regulate the overall activity and functional outputs of these systems, but how these interactions enter into the regulatory decisions of individual cells and form the basis for the emergent properties of these cellular networks remains poorly understood. We combine quantitative biological modeling with experiments using wild and genetically engineered bacterial strains to dissect the parameters that regulate microbial activity within single cells, spatially structured populations, and well-mixed microcosms. I will discuss our recent work on the structure of multispecies interaction networks, pattern formation, and engineering synthetic microbial communities.
Flyer James_Boedicker_20161027.pdf

CANCELLED - Guiding Antibody Evolution via Programmable Antigen Environments

Date Thursday October 20, 2016 at 4:00 PM
Location A2-342 MDCC, Moss Auditorium, Marion Davies Children Clinic
Speaker Shenshen Wang, Ph.D., Assistant Professor (Theoretical/Computational Biophysics), Department of Physics & Astronomy, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract ABSTRACT: Antibodies, soluble forms of B cell receptors, attain increasing affinity for encountered antigens by alternating between somatic mutation and competitive selection in a cyclic fashion, a phenomenon known as affinity maturation. But this Darwinian process becomes ineffective in facing highly mutable complex pathogens, notably HIV, which have evolved various tactics to escape and divert immune responses. In this talk, I will describe the challenges in inducing broadly neutralizing antibodies by vaccination, present an agent-based computational model of affinity maturation driven by multiple antigens with complex epitopes, and discuss novel immunization strategies that can potentially focus antibody response onto the targeted viral vulnerability which test favorably in mice. If time allows, our recent attempts and surprises will be briefly mentioned.
Flyer Shenshen_Wang_20161020_cancelled.pdf

Genomic Prediction of Quantitative Traits

Date Thursday October 13, 2016 at 4:00 PM
Location 53-105 CHS Center for the Health Sciences
Speaker Steve Hsu, Ph.D., Professor of Theoretical Physics and Vice-President for Research, Michigan State University
Sponsoring Dept UCLA Biomathematics
Abstract I discuss the application of Compressed Sensing (L1-penalized optimization or LASSO) to genomic prediction. I show that matrices comprised of human genomes are good compressed sensors, and that LASSO applied to genomic prediction exhibits a phase transition as the sample size is varied. When the sample size crosses the phase boundary complete identification of the subspace of causal variants is possible. For typical traits of interest (e.g., with heritability ~ 0.5), the phase boundary occurs at N ~ 30s, where s (sparsity) is the number of causal variants. I give some estimates of sparsity associated with complex traits such as height and cognitive ability, which suggests ~ 10k. In practical terms, these results imply that powerful genomic prediction will be possible for many complex traits once ~ 1 million genotypes are available for analysis.
Flyer Steve_Hsu_20161013.pdf
Video Seminar Video
Slides Seminar Slides

Mathematical Oncology at City of Hope

Date Thursday October 06, 2016 at 4:00 PM
Location A2-342 MDCC, Moss Auditorium, Marion Davies Children Clinic
Speaker Russell C. Rockne, Ph.D., Assistant Professor, Director of Mathematical Oncology, Dept. of Information Sciences, Beckman Research Institute, City of Hope
Sponsoring Dept UCLA Biomathematics
Abstract In this seminar, we will highlight mathematical models that have been used to provide predictions and insight into clinical challenges in oncology. We will also present a high-level survey of the division of mathematical oncology at City of Hope, and describe a roadmap of how we intend to achieve our mission of translating mathematics-based research into clinical care.
Flyer Russell_Rockne_20161006.pdf

Single-Nucleotide Analysis of RNA-Seq: Methodologies and Applications

Date Thursday September 22, 2016 at 4:00 PM
Location A2-342 MDCC, Moss Auditorium, Marion Davies Children Clinic
Speaker Xinshu (Grace) Xiao, Ph.D., Department of Integrative Biology and Physiology, University of California, Los Angeles
Sponsoring Dept UCLA Biomathematics
Abstract High-throughput sequencing of RNA (RNA-Seq) is becoming widely applied in the biomedical and clinical realms. This technology generates an enormous amount of data that captures the gene expression landscape of the genome globally. Many aspects of gene regulation can be examined using RNA-Seq and related techniques. Here, we will focus on single-nucleotide analysis and address the challenges and opportunities in this area. We will introduce newly developed methods ranging from read mapping to segregation of genetic variants and RNA editing sites. Applications of these methods to study RNA editing in a large number of human samples will be presented. RNA editing is an important post-transcriptional mechanism where a single nucleotide in the RNA can be modified and converted into a different nucleotide, thus diversifying the expression of the genome. These applications highlight the value of the methods in revealing novel evolutionary and regulatory patterns of RNA editing.
Flyer Xinshu_Grace_Xiao_20160922.pdf

Learning Clinical Outcomes from Massive Observational Data

Date Wednesday June 22, 2016 at 10:00 AM
Location 12-407 MDCC (Marion Davies Children Center), Friedman Board Room
Speaker Trevor Shaddox, Doctoral Graduate Student, Department of Biomathematics, Friedman Board Room
Sponsoring Dept UCLA Biomathematics
Abstract Emerging national patient claims and electronic health record databases offer a rich frontier for learning about treatment effectiveness and clinical decision making. However, these resources present statistical and computational challenges commensurate with their promise, requiring innovative approaches for practically and efficiently extracting meaningful results. In this defense, I seek to address some of these challenges. First, I present a hierarchical model for learning about the relationship between treatments and multiple related adverse outcomes simultaneously, showing that this approach can reduce bias in relative risk estimates. Second, I develop a novel minorization-maximization (MM) algorithm for uncoupling the sequential Newton steps that arise within the state of the art model fitting procedure for the conditional models popular for observational studies, allowing faster, parallelized model fitting. Third, I develop a birth-death model for treatment trajectories among patients with diabetes mellitus type II. In each topic, I discuss applications to observational healthcare datasets, demonstrating how these methods work at scale.
Flyer Trevor_Shaddox_defense_seminar.pdf

Projection Algorithms for Large Scale Optimization and genomic data analysis

Date Friday June 10, 2016 at 10:00 AM
Location Gonda Center conference room 1357
Speaker Kevin Keys, Doctoral Graduate Student, Department of Biomathematics, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract The advent of the Big Data era has spawned intense interest in scalable optimization methods. Traditional approaches such as Newton’s method fall apart whenever the features outnumber the examples in a data set. Consequently, researchers have intensely developed first-order methods that rely only on gradients and subgradients of a cost function. In this dissertation we focus on projected gradient methods for large-scale constrained optimization. We develop a particular case of a proximal gradient method called the proximal distance algorithm that combines the classical penalty method of constrained minimization with distance majorization. To optimize the loss function f(x) over a constraint set C, the proximal distance principle mandates minimizing the penalized loss f(x) + t dist(x,C)2 and following the solution x to its limit as t tends to infinity. At each iteration dist(x,C)2 is majorized by || x - ΠC(xk) ||2, where ΠC (xk) denotes the projection of the current iterate xk onto C. The minimum of f (x) + t || x - ΠC (xk) ||2 is given by the proximal map prox(1/t) f [ΠC(xk)]. Since many projections and proximal maps are known in analytic or computable form, the proximal distance algorithm provides a scalable computational framework for a variety of constraints. For the particular case of sparse linear regression, we implement a projected gradient algorithm known as iterative hard thresholding (IHT) for genome-wide association studies. A genome-wide association study (GWAS) correlates marker variation with trait variation in a sample of individuals genotyped at a multitude of SNPs (single nucleotide polymorphisms) spanning the genome. The massive amount of data produced in these studies present unique computational challenges. Penalized regression with LASSO or MCP penalties is capable of selecting a handful of associated SNPs from millions of potential SNPs. Unfortunately, type I and type II errors can complicate model selection and obscure the genetic underpinning of a trait. Our parallel implementation of IHT accommodates SNP genotype compression and exploits both multicore CPUs and massively parallel GPUs. This allows statistical geneticists to leverage desktop workstations and to eschew expensive supercomputing resources. We evaluate IHT performance on both simulated and real GWAS data and conclude that it reduces type I and type II errors while maintaining compute speeds competitive with LASSO and MCP.
Flyer Kevin_Keys_defense_seminar.pdf

Mathematical Modeling of Chromosomal Transport Mechanisms

Date Thursday May 19, 2016 at 4:00 PM
Location 53-105 CHS (Center for the Health Sciences)
Speaker Blerta Shtylla, Ph.D., Assistant Professor, Department of Mathematics, Pomona College
Sponsoring Dept UCLA Biomathematics
Abstract The generation of directed movement of cellular components frequently requires the rectification of Brownian motion. In this talk, we discuss mathematical models that track bias generation by nano machine constructs that operate during cell division. The efficient operation of these dynamic constructs requires specific interactions with dynamic bio polymers. We use first passage techniques to derive mesoscale properties for these motor constructs using microscale rates and reactions.
Flyer Blerta_Shtylla_seminar_20160519.pdf

The Ecology and Evolution of Social Aggregations: Case Study Dictyostelium Discoideum

Date Thursday May 05, 2016 at 4:00 PM
Location 13-105 Center for the Health Sciences (CHS)
Speaker Corina Tarnita, Ph.D., Princeton University
Sponsoring Dept UCLA Biomathematics
Abstract ABSTRACT: Cooperation, in which individuals provide benefits to others at a cost to themselves, has been studied extensively, and mechanisms have been proposed for its persistence in the face of free-riders. Often however, especially in microbes, these studies focus on one fitness component, with little information about or attention to the ecological context, which can lead to paradoxical findings. I will discuss such an example in the slime mold Dictyostelium discoideum whose life cycle includes both a single cellular and a multicellular stage, and I will propose a broader ecological framework in which multiple life history tradeoffs arise collectively in response to characteristics of the environment. I will argue that this multidimensionality can resolve existing inconsistencies regarding the social behavior and I will conclude that the complexities of social behavior in general and multicellularity in particular can only be understood in the appropriate ecological and life history context
Flyer Corina_Tarnita_20160505_updated.pdf

Computation of Transition States and its Applications in Biology

Date Thursday January 14, 2016 at 4:00 PM
Location 13-105 Center for the Health Sciences (CHS)
Speaker Lei Zhang, Ph.D., Beijing International Center for Mathematical Research and Center for Quantitative Biology, Peking University
Sponsoring Dept UCLA Biomathematics
Abstract ABSTRACT: The dynamics of complex biological systems is often driven by multiscale, rare but important events. In this talk, I will first introduce the numerical methods for computing transition states, in particular, the Optimization-based Shrinking Dimer (OSD) method we recently proposed. Then I will give two applications of rare events and transition states in biology, including boundary sharpening in zebrafish hindbrain and neuroblast delamination in Drosophila. The joint work with Qiang Du (Columbia), Qing Nie (UC Irvine), Yan Yan (HKUST).
Flyer Lei_Zhang_20161014.pdf

Predicting and Designing Self-Assembling Systems

Date Thursday January 07, 2016 at 4:00 PM
Location 13-105 CHS, Center for the Health Sciences
Speaker Martin Nilsson Jacobi, Ph.D., Professor, Department of Energy and Environment, Chalmers University, Sweden
Sponsoring Dept UCLA Biomathematics
Abstract I will present a theoretical method for predicting and designing self-assembling systems. I will demonstrate that, by using analytic solution to a statistical mechanics model (the spherical spin model), it is possible to predict several interesting pattern formation phenomena in systems of point particles with isotropic interactions. Further, using the same basic idea, I will show that it is possible to design the interactions so that the particles self-assemble into target lattice structures. Last, I will demonstrate how these ideas can also be used to predict and design self-assembly of so called patchy colloids, i.e. nano-particles with spots and stripes on their surface.
Flyer Martin_Nilsson_Jacobi_20160197_seminar.pdf

Real Cardiovascular Problems, Resolved by Mathematics and Realized by Knitting

Date Thursday December 10, 2015 at 4:00 PM
Location A2-342 MDCC, Moss Auditorium, Marion Davies Children's Clinic
Speaker Torbjörn Lundh, Ph.D., Professor, Department of Mathematical Sciences, Chalmers University of Technology, on a sabbatical at Dept of Surgery @ Stanford
Sponsoring Dept UCLA Biomathematics
Abstract We will present four different cardiovascular problems that all were stated by vascular surgeons. We addressed them using highly diverse mathematical "approaches" and suggest "solutions". Various knitting techniques was used to make textile prototypes of three of the solutions. The fourth one is a very old problem about wound healing and unfortunately still a bit open.
Flyer Torbjorn_Lundh_20151210.pdf

Anomalous Diffusion and Random Encounters in Living Systems

Date Thursday December 03, 2015 at 4:00 PM
Location A2-342 MDCC, Moss Auditorium, Marion Davies Children's Clinic
Speaker Scott A. McKinley, Ph.D., Assistant Professor, Department of Mathematics, Tulane University
Sponsoring Dept UCLA Biomathematics
Abstract The last twenty years have seen a revolution in tracking data of biological agents across unprecedented spatial and temporal scales. An important observation from these studies is that path trajectories of living organisms can appear random, but are often poorly described by classical Brownian motion. The analysis of this data can be controversial because practitioners tend to rely on summary statistics that can be produced by multiple, distinct stochastic process models. Furthermore, these summary statistics inappropriately compress the data, destroying details of non-Brownian characteristics that contain vital clues to mechanisms of transport and interaction. In this talk, I will survey the mathematical and statistical challenges that have arisen from my work on the movement of foreign agents, including viruses and synthetic microparticle probes, in human mucus. My collaborators and I have demonstrated that the behavior of individual particles is well-described by the integrated Generalized Langevin Equation, a Gaussian process that features tunable autocorrelation in time. Physicists have postulated that the memory in particle paths is related to certain viscoelastic features of the fluid environment. I will detail the stochastic PDE framework necessary to probe this hypothesis and detail some successes (and failures!) in describing population scale dynamics in such a way to predict the rate at which these agents penetrate the human body's first line of defense.
Flyer Scott_McKinley_20151203.pdf

Connecting Experiments, Data Standards, and 3-D Simulations for Multicellular Systems Biology

Date Thursday November 19, 2015 at 4:00 PM
Location A2-342 MDCC, Moss Auditorium, Marion Davies Children's Clinic
Speaker Paul Macklin, Ph.D., Assistant Professor, Center for Applied Molecular Medicine, University of Southern California
Sponsoring Dept UCLA Biomathematics
Abstract ABSTRACT: New experimental and imaging techniques have created a deluge of molecular and multicellular data, but these data have not necessarily improved our understanding of multicellular biological systems. One solution is to use mechanistic 3-D simulators for hypothesis testing, engineering, and therapeutic planning. In this talk, we will present our work to create a suite of open source tools and data standards for multicellular systems biology and medicine. We will show parameter identification tools that extract cell birth and death parameters from high-content screening experiments. We store these and other biophysical parameters in a library of model-independent digital cell lines using MultiCellDS (multicellular data standard). We will present BioFVM (finite volume method for biology) and PhysiCell (physics-based cell simulator), which jointly can simulate millions of cells in 3-D tissues. The C++ codes use OpenMP to enable large simulations (at least 5 million cells, 10 diffusing substrates, 1 million voxels) on quad-core desktops and individual supercomputer nodes, or large parameter investigations distributed across high performance clusters. We will close with applications to bioengineering experiments and patient-tailored cancer simulations.
Flyer Paul_Macklin_20151119.pdf

RNA as a Linear Polymer, but a Branched Genome

Date Thursday November 12, 2015 at 4:00 PM
Location A2-342 MDCC, Moss Auditorium, Marion Davies Children's Clinic
Speaker William M. Gelbart, Ph.D., Professor, Department Chemistry & Biochemistry, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract We learn in school that the genetic material of life is DNA. But the genome of most viruses is single-stranded (ss) RNA, as opposed to double-stranded (ds) DNA. And, even though ssRNA is strictly a linear polymer -- involving a chain of covalently-linked nucleotides -- it behaves effectively as a highly branched polymer, because of the large extent of self-complementarity (base-pairing between distant nucleotides along the chain). In my talk I discuss how we characterize and quantify the “branchedness” of long RNA molecules, and its role in determining the physical properties of virus-like particles and the infectivity of viruses.
Flyer William_Gelbart_20151112.pdf

Multi-Scale Dynamics of Calcium Signaling in Cardiac Myocytes

Date Thursday November 05, 2015 at 4:00 PM
Location A2-342 MDCC, Moss Auditorium, Marion Davies Children's Clinic
Speaker Zhilin Qu, Ph.D., Professor, Department of Medicine, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract Intracellular calcium (Ca) signaling is a ubiquitous signaling process in biology, regulating biological functions from life to death. Besides biological signal transduction, Ca is required for muscle contraction and plays a key role in generating both normal and abnormal cardiac rhythms. Our group uses mathematical modeling, computer simulation, experiments, as well as nonlinear dynamics to investigate the dynamics of Ca signaling in cardiac myocytes. In this talk, I will present the novel mathematical theories we developed recently for: 1) termination and duration of Ca sparks; 2) transition from random Ca sparks to waves and whole-cell oscillations; and 3) genesis of arrhythmogenic Ca alternans. I will present experimental results in debate for each topic, and show how our theories can reconcile and unify these seemingly contradictory experimental observations.
Flyer Zhilin_Qu_20151105.pdf

Minorization-Maximization Algorithms for Variance Components Models

Date Thursday October 29, 2015 at 4:00 PM
Location A2-342 MDCC, Moss Auditorium, Marion Davies Children's Clinic
Speaker Hua Zhou, Ph.D., Associate Professor, Department of Biostatistics, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract Variance components estimation and mixed model analysis are central themes in statistics with applications in numerous scientific disciplines. Despite the best efforts of generations of statisticians and numerical analysts, maximum likelihood estimation and restricted maximum likelihood estimation of variance component models remain numerically challenging. In this talk, we present a novel iterative algorithm for variance components estimation based on the minorization-maximization (MM) principle. MM algorithm is trivial to implement and competitive on large data problems. The algorithm readily extends to more complicated problems such as linear mixed models, multivariate response models possibly with missing data, maximum a posteriori estimation, penalized estimation, and generalized estimating equations (GEE). We demonstrate, both numerically and theoretically, that it converges faster than the classical EM algorithm when the number of variance components is greater than two. This talk is accessible to graduate students especially those taking BIOMATH 210.
Flyer Hua_Zhou_20151029.pdf

Benefits and Costs of Mutational Robustness in RNA Viruses

Date Thursday October 22, 2015 at 4:00 PM
Location A2-342 MDCC, Moss Auditorium, Marion Davies Children's Clinic
Speaker Simone Bianco, Ph.D., Research Staff Member, IBM Almaden Research Center, San Jose, California
Sponsoring Dept UCLA Biomathematics
Abstract The accumulation of mutations in RNA viruses is thought to facilitate rapid adaptation to changes in the environment. However, most mutations have deleterious effects. Thus, tolerance to mutations should determine the nature and extent of genetic diversity in the population. I will present a combination of population genetics theory, computer simulation, and experimental evolution to examine the advantages and disadvantages of tolerance to mutations, also known as mutational robustness. Our findings may inform therapeutic strategies that cause extinction of otherwise robust viral populations.
Flyer Simone_Bianco_20151022.pdf

Admixture, Genography, and Clustering

Date Thursday October 15, 2015 at 4:00 PM
Location A2-342 MDCC, Moss Auditorium, Marion Davies Children's Clinic
Speaker Ken Lange, Ph.D., Departments of Biomathematics and Human Genetics, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract This talk will deal with three related topics of ancestry inference from genetic data. Topic one concerns estimation of admixture proportions for people of mixed ethnicity. Population stratication has long been recognized as a potential confounding factor in genetic association studies. Estimated ancestries, derived from multi-locus genotype data, can be used as covariates to correct for population stratication. Topic two summarizes how one can locate the geographic origin of individuals based on their genetic backgrounds. SNPs (single nucleotide polymorphisms) vary widely in informativeness, allele frequencies change nonlinearly with geography, and reliable localization requires evidence to be integrated across a multitude of SNPs. These problems become more acute for individuals of mixed ancestry. Topic three summarizes a new convex clustering method that delivers an entire clustering path. When applied to genetic data, these paths recapitulate the divergence of human populations.
Flyer Ken_Lange_20151015A.pdf

Terasaki Ramps in the Endoplasmic Reticulum: Structure, Function and Formation

Date Thursday October 08, 2015 at 4:00 PM
Location A2-342 CHS, Moss Auditorium, Marion Davies Children's Clinic
Speaker Greg Huber, Ph.D., Kavli Institute for Theoretical Physics, Department of Physics, University of California, Santa Barbara
Sponsoring Dept UCLA Biomathematics
Abstract The endoplasmic reticulum (ER) has long been considered an exceedingly important and complex organelle in eukaryotes. It is a membrane structure, part folded lamellae, part tubular network, that both envelopes the nucleus and threads its way outward, all the way to the cell’s periphery. Despite the elegant mechanics of bilayer membranes offered by the work of Helfrich and Canham, as far as the ER is concerned, theory has mostly sat on the sidelines. However, refined imaging of the ER has recently revealed beautiful and subtle geometrical forms – simple geometries, from the mathematical point of view – which some have called a “parking garage for ribosomes.” I’ll review the discovery and physics of Terasaki ramps and discuss their relation to cell-biological questions, such as ER and nuclear-membrane re-organization during mitosis. Rather than being a footnote in a textbook on differential geometry, these structures suggest answers to a number of the ER’s structure-function problems
Flyer Greg_Huber_20151008.pdf

Neural Codes, Place Fields, and Convexity

Date Thursday October 01, 2015 at 4:00 PM
Location A2-342 CHS, Moss Auditorium, Marion Davies Children's Clinic
Speaker Nora Youngs, Ph.D., Postdoctoral Fellow, Department of Mathematics, Harvey Mudd College
Sponsoring Dept UCLA Biomathematics
Abstract Navigation is one of the most important functions of the brain. This year, the Nobel Prize in Medicine and Physiology was awarded for the discovery of place cells and grid cells, neurons which form vital pieces of the navigation system. Though the external observed correspondence of these neurons to 2D receptive fields has been carefully recorded and proven, the animal itself navigates the world without access to these maps. An important problem confronted by the brain is to infer what properties of a stimulus space can - in principle - be extracted from the stimulus space. This motivates us to define the neural ring and a related neural ideal, algebraic objects that encode the combinatorial data of a neural code. We find that these objects can be expressed in a “canonical form’’ that directly translates to a minimal description of the receptive field structure intrinsic to the neural code, and present an algorithm to compute this canonical form. We also find that topological information about the stimulus space can be directly extracted from the neural ideal.
Flyer Nora_Youngs_20151001.pdf

Coupled Reaction-Diffusion Models with Degenerate Sources

Date Thursday April 02, 2015 at 4:00 PM
Location 53-105 Center for the Health Sciences (CHS)
Speaker Jonathan Wylie, Ph.D., Professor, Department of Mathematics, City University of Hong Kong
Sponsoring Dept UCLA Biomathematics
Abstract We consider a general system of coupled nonlinear diffusion equations that are characterized by having degenerate source terms and thereby not having isolated rest states. Such equations arise naturally in the study of ion propagation through biological cells and fluid transport through porous media with evaporation and condensation. Using a general form of physically relevant source terms, we derive conditions that are required to trigger traveling waves when a stable uniform steady-state solution is perturbed by a highly localized disturbance. We also discuss the implications for biological systems.
Flyer Jonathan_Wylie_20150402.pdf

Developing and Integrating Computer-Aided Diagnostic Tools into Clinical Medicine

Date Tuesday March 31, 2015 at 2:00 PM
Location MacDonald Research Lab (MRL) 1-441
Speaker Wesley Kerr, Doctoral Graduate Student, Department of Biomathematics, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract This thesis is comprised of two parts (1) development of unimodal and multimodal computer-aided diagnostic tools (CADTs) for seizure disorder and (2) a novel method for optimization of hyperparameters in machine learning models. The aims of CADTs are to address key challenges in the diagnosis and treatment of seizure disorder, including reducing the time to an accurate diagnosis, improving the sensitivity and specificity of diagnostic neuroimaging, and the understanding of the diagnostic value of interictal scalp electroencephalography. Our novel method for optimizing hyperparameters has the potential to slightly improve the accuracy of machine learning models, while substantially increasing the interpretability of learned estimates and reducing computational cost.
Doctoral Committee: Mark S. Cohen, Ph.D., Henry Huang, D.Sc., Elliot M. Landaw, M.D., Ph.D., Marc A. Suchard, M.D., Ph.D., John M. Stern, M.D.
Flyer Wesley_Kerr_defense_seminar_20150331.pdf

Data–Driven Approaches to Protein Biophysics

Date Friday February 20, 2015 at 4:00 PM
Location 132 Neuroscience Research Building Auditorium
Speaker Julie Mitchell, Ph.D., Professor, Department of Biochemistry and Mathematics, University of Wisconsin-Madison
Sponsoring Dept UCLA Biomathematics
Abstract ABSTRACT: The Big Data era offers new avenues in molecular modeling through the integration of classical physics and chemistry with data-driven computation. Recasting biophysical questions in terms of classification gives rise to robust predictive models and avoids many challenges in working with non-convex energy functions, using knowledge derived from ab initio force fields and simulation as features in predictive models of molecular association. Several successful examples of this strategy in developing models for alanine mutagenesis, allosteric communication and DNA binding will be presented. Recent applications also highlight the value of integrating data on molecular evolution into physical studies of protein interaction.
Host: Maria D’Orsogna, Ph.D.
Flyer Julie_Mitchell_20150220.pdf

Stochastic Processes at Single-Molecule and Single-Cell Levels

Date Friday January 16, 2015 at 4:00 PM
Location 53-105 Center for the Health Sciences (CHS)
Speaker Hao Ge, Sc.D., Associate Professor, Beijing International Center for Mathematical Research, Biodynamic Optical Imaging Center, Peking Universit
Sponsoring Dept UCLA Biomathematics
Abstract Due to the advance of single-molecule techniques, stochastic phenomena in chemistry and biology have been widely observed, which promotes the rapid development of stochastic modeling. I will discuss several stochastic processes in single-molecule enzyme kinetics, transcriptional burst and toggle switch. The stochastic modeling not only can explain some unexpected law from the trajectory perspective, but also can help uncover certain molecular mechanism and carefully analyze the effect of noise within gene regulation.
Host: Tom Chou, Ph.D.
Flyer 20150116_Hao_Ge.pdf

Theoretical Approaches towards HIV Vaccine Designs and Incidence Assessment

Date Thursday December 11, 2014 at 4:00 PM
Location 13-105 Center for the Health Sciences (CHS)
Speaker Ha Youn Lee, Ph.D., Associate Professor, Department of Molecular Microbiology and Immunology, Keck School of Medicine, Univ. of Southern California
Sponsoring Dept UCLA Department of Biomathematics
Abstract The world is facing a serious global pandemic of HIV/AIDS, with more than 35 million people infected. Controlling and eventually eradicating this unprecedented pandemic will require a better surveillance tool and improved vaccine design. An effective HIV vaccine must induce immune responses that recognize as many HIV strains as possible in order to better protect against the rapidly-mutating virus. The surface morphology of peptide-MHC (pMHC) complexes is one of the key factors controlling the breadth of reactive T cell population. Here we present our computational design for predicting pMHC surface morphology, which assembles homology-based models and all-atom molecular dynamics simulations. Our method shows high precision, with root mean square deviation=1.58 Å over a 17 pMHC test set. A blind test is performed on three peptides with undetermined structure, and high resolution X-ray crystallography data has corroborated our predictions. Once a functional HIV vaccine is implemented, monitoring HIV incidence, the number of newly infected people, is necessary to evaluate its efficacy. There is no universal standard for assays measuring HIV incidence since conventional approaches rely on variable and inaccurate HIV-specific antibody responses. Our group has focused on developing HIV genomic incidence assays utilizing signatures embedded in an individual’s HIV sequence population. Since HIV sequences are evolving throughout the course of infection, we can quantify the amount of evolution as a fingerprint of infection duration. We have produced over 18,000 HIV envelope gene segments by combining high-throughput next-generation sequencing and a signal-masking bioinformatics pipeline. Two biomarkers we created successfully related sequence similarity to the infection stage, distinguishing between recent and chronic infections with over 95% accuracy. Furthermore, mathematical modeling has allowed us to extend our accuracy in determining infection duration for recently infected individuals. Taken together, analytical approaches are becoming indispensable components to medical research.
Host: Tom Chou, Ph.D.

To receive e-mail seminar notices, contact David Tomita (
Flyer 20141211_Ha_Youn_Lee.pdf

Biological Regulation of Calcification and Field Theory for Inverse Problems

Date Thursday December 04, 2014 at 4:00 PM
Location 23-105 Center for the Health Sciences (CHS)
Speaker Joshua Chang, Ph.D., Postdoctoral Fellow, Mathematical Biosciences Institute, Columbus, Ohio
Sponsoring Dept UCLA Biomathematics
Abstract In this two-part talk, I will present a model for the regulation of precipitation of calcium phosphate species in biological tissues. Calcium is an important ion for both structural support and biochemical signaling in vertebrates. As a result, it is necessarily maintained at high concentrations in fluids - at levels where precipitation is favored. Yet, such precipitation, when it occurs in an uncontrolled manner, is harmful. Using concepts from classical nucleation theory, I will discuss how biological organisms can regulate this high calcium concentration. Nucleation and crystallization problems such as this one are often studied through the use of atomic force microscopy (AFM). AFM and related techniques are associated with inverse problems of Brownian motion. In the second part of my talk, I will discuss the inverse problem of potential energy reconstruction for random walkers under non-constant diffusivity. I will present a self-contained, nonparametric, regularized method based on Bayesian inference under which a path integral is used for uncertainty quantification. Under this method, regularization parameters are determined through optimization of an eigenvalue problem for a trace-class operator.
Host: Tom Chou, Ph.D.
Flyer 20141204_Joshua_Chang.pdf

Feedback, Lineages and Vascular Tumor Growth

Date Thursday November 20, 2014 at 4:00 PM
Location 13-105 Center for the Health Sciences (CHS)
Speaker John Lowengrub, Ph.D., Professor, Department of Mathematics, UC Irvine
Sponsoring Dept UCLA Biomathematcis
Abstract ABSTRACT: Cancer arises when the carefully regulated balance of cell proliferation and programmed cell death (apoptosis) that ordinarily exists in normal homeostatic tissues is disrupted. Cancer cells are assumed to acquire a common set of traits. However, not all the cells in a tumor seem to matter equally and tumor cells progress through lineage stages regulated by feedback pathways. It is known that the microenvironment plays an important role in this regulation. Secreted factors by vascular endothelial cells (ECs) have been found to support and maintain cancer stem cells (CSCs) and it has even been observed that CSCs can transdifferentiate into ECs. It has been hypothesized that transdifferentiated ECs may contribute to tumor vascularization via vasculogenesis. However, these processes are not well understood and mathematical modeling can provide insight on the underlying biology. We use a hybrid continuum-discrete multispecies mathematical model to simulate numerically the three-dimensional spatiotemporal dynamics of hierarchically-structured, vascularized solid tumors. Tumor cells and substrate species are treated as continuum, while vessels are treated as discrete quantities. We account for protein factors secreted by tumor cells and ECs that affect angiogenesis, tumor cell self-renewal, differentiation and transdifferentiation, and proliferation pathways. By testing different combinations, our models reveal the effects of feedback regulation on tumor size, invasiveness as well as on the heterogeneous distribution of cells within the tumor and the structure of the vascular network. Consistent with experimental observations, positive feedback from the ECs to the CSCs creates perivascular niches and increases the CSC fractions in the tumor as well as the tumor sizes and the amount of functional vasculature. Intratumoral vasculogenesis is found to result from transdifferentiation of CSCs into ECs, thereby increasing tumor sizes further. Negative regulation by ECs on CSCs transdifferentation is paradoxically found to increase tumor sizes, CSC fractions and vasculogenesis. The close interactions between tumor cells and the vascular network present opportunities and challenges for therapeutic intervention.
Host: Mary Sehl, M.D.

To receive e-mail seminar notices, contact David Tomita (
Flyer John_Lowengrub_20141120.pdf

The Interplay of Antibiotic Inhibition and Inactivation Promotes Microbial Diversity

Date Thursday November 13, 2014 at 4:00 PM
Location 13-105 Center for the Health Sciences (CHS)
Speaker Kalin Vetsigian, Ph.D., Asst. Prof., Dept. of Bacteriology & the Systems Biology Theme of Wisconsin Institute for Discovery, Univ. of Wisconsin-Madison
Sponsoring Dept UCLA Biomathematics
Abstract ABSTRACT: Many natural environments harbor microbial communities exhibiting a great diversity of antibiotic production and resistance capabilities. This raises the question of whether inhibitory antibiotic interactions stabilize or disrupt diversity. Theoretical studies and microcosm experiments have shown that inhibitory interactions can promote diversity through relationships of cyclic dominance (e.g. paper-scissor-rock games) they generate. However, such theories rely on preservation of spatial community structure and do not predict stable diversity under realistic assumptions of microbial dispersal. In this work we demonstrate that antibiotic inactivation, in which one species alleviates the antibiotic inhibition of another species, is widespread among antibiotic producing soil bacteria and that this type of higher-order interaction enables robust diversity maintenance of species with different antibiotic production and resistance capabilities even at high levels of microbial dispersal. More generally, this illustrates that competitive pairwise interactions can promote diversity when higher order effects are taken into account, which is likely of importance in many ecosystems.
Host: Van Savage, Ph.D.

To receive e-mail seminar notices, contact David Tomita (
Flyer Kalin_Vetsigian_20141113.pdf

Unexpected Responses of Disease to Global Change

Date Thursday November 06, 2014 at 4:00 PM
Location 13-105 Center for the Health Sciences (CHS)
Speaker Erin Mordecai, Ph.D., Postdoctoral Fellow, Biology Dept., Univ. of N. Carolina, and N. Carolina State Univ. (Starting Asst. Prof. at Stanford in 2015)
Sponsoring Dept UCLA Biomathematics
Abstract ABSTRACT: With the threat of changing climate, species invasions, shifts in land use, and other anthropogenic changes, ecologists are increasingly concerned about the emergence and spread of infectious diseases. The common assumption is that environmental changes will facilitate disease spread and increase the risk to humans and species of agricultural and conservation concern. Yet most ecological processes are nonlinear, and the response of infectious diseases to environmental change is no exception. In this talk, I will explore how nonlinearities in disease transmission lead to unexpected responses of disease to environmental change in two systems: (1) a fungal pathogen called Black Fingers of Death that spills over from invasive cheatgrass to native grass species in the western U.S., and (2) changes in human malaria risk in response to temperature. To understand the importance of nonlinearity in these systems, I use mathematical models fit to empirical data. In both cases the field-parameterized models show, counterintuitively, that environmental change does not necessarily lead to negative diseasemediated outcomes. In fact, the fungal pathogen is predicted to benefit the native grass species in competition with invasive cheatgrass, and warm temperatures are expected to decrease malaria transmission in currently heavily-infected areas. These surprising results underscore the importance of integrating models and data to predict responses of disease to environmental change in nature.
Host: Van Savage, Ph.D.

To receive e-mail seminar notices, contact David Tomita (
Flyer Erin_Mordecai_20141106.pdf

Lotka’s Dilemma, Search Strategies, and the Pace of Ecological Interactions

Date Thursday October 30, 2014 at 4:00 PM
Location 13-105 Center for the Health Sciences (CHS)
Speaker Andrew Hein, Ph.D., Postdoctoral Fellow, Dept. of Ecology & Evolutionary Biology, Princeton University
Sponsoring Dept UCLA Biomathematics
Abstract ABSTRACT: In the 1920’s, pioneering ecologist Alfred Lotka and others introduced the models that have formed the basis of modern theoretical ecology, and the theory of species interactions (e.g., consumer-resource interactions) in particular. The descendants of these models are used in fields from evolutionary ecology to biogeochemistry. In this talk, I’ll revisit the assumptions of such models and show, as Lotka warned, that they do not allow for the sorts of complex behavioral responses to resources (e.g., active search behavior) that abound in real ecological systems. I’ll describe how active search behavior can be incorporated, and demonstrate that it can change the rates of ecological interactions in quantitative and qualitative ways.
Host: Van Savage, Ph.D.

To receive e-mail seminar notices, contact David Tomita (
Flyer Andrew_Hein_20141030.pdf

Integrative Statistical Approaches to Find Causal Variants in Post-GWAS Studies

Date Thursday October 23, 2014 at 4:00 PM
Location 13-105 Center for the Health Sciences (CHS)
Speaker Bogdan Pasaniuc, Ph.D., Assistant Professor, Dept. of Pathology & Laboratory Medicine, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract Genome-wide association studies (GWAS) have successfully identified numerous regions in the genome that harbor genetic variants that increase risk for various complex traits and diseases. However, it is generally the case that GWAS risk variants are not themselves causally affecting the trait, but rather, are correlated to the true causal variant through linkage disequilibrium (LD). Plausible causal variants are identified in fine-mapping studies through targeted sequencing followed by prioritization of variants for functional validation. In this work, we propose methods that leverage two sources of independent information, the association strength and genomic functional location, to prioritize causal variants. We demonstrate in simulations and empirical data that our approach reduces the number of SNPs that need to be selected for follow-up to identify the true causal variants at GWAS risk loci.
Host: Janet Sinsheimer, Ph.D.
To receive e-mail seminar notices, contact David Tomita (
Flyer 20141023_Bogdan_Pasaniuc.pdf

Assessing the Transmission Potential of Spillover Infections

Date Thursday October 16, 2014 at 4:00 PM
Location 13-105 Center for the Health Sciences (CHS)
Speaker Adam Kucharski, Ph.D., Research Fellow, Department of Infectious Disease Epidemiology, London School of Hygiene & Tropical Medicine
Sponsoring Dept UCLA Biomathematics
Abstract ABSTRACT: Obtaining good estimates of transmission is crucial for effective surveillance and control of infectious diseases. However, when an infection transmits inefficiently between humans, estimates often have to be made using case data from a limited number of small outbreaks. I will talk about some of our recent work on the problem, which combines social contact surveys with multi-type branching processes to estimate the transmission potential of new infections. As well as covering the theoretical challenges involved, I will discuss how the work can be used to understand spillover infections such as monkeypox, influenza A(H5N1) and MERS-CoV.
Host: James Lloyd-Smith, Ph.D.

To receive e-mail seminar notices, contact David Tomita (
Flyer Adam_Kucharski_20141016.pdf

Neural Bursts, Averaging and Canards

Date Tuesday September 16, 2014 at 4:00 PM
Location 13-105 Center for the Health Sciences (CHS)
Speaker Martin Wechselberger, Ph.D., Associate Professor, School of Mathematics and Statistics, University of Sydney
Sponsoring Dept UCLA Biomathematics
Abstract ABSTRACT: Bursting behavior in neurons has been the focus of significant theoretical attention due to both its mathematical complexity and its central role in driving repetitive actions such as respiration and hormone release. A wide variety of forms of bursting that arise in fast-slow single-neuron models are well understood based on fast-slow decomposition, identification of fast subsystem bifurcation structures, and averaging, and these methods also can be used to explain transitions between quiescence, bursting, and tonic spiking in single neurons. Transitions between such activity patterns in neuronal network models, however, are much less well understood. In this talk, we identify generic bifurcation scenarios corresponding to transitions between bursting and tonic spiking solutions in a model for a coupled pair of burst-capable neurons, and we elucidate the central role of folded singularities in these scenarios. The folded singularities in our work arise in the context of fast-slow averaging (see, e.g., [Cymbalyuk, Shilnikov (2005), J. Comp. Neurosci. 18, 255-263]) and hence our results link with the study of torus canards, a recently identified class of solutions featuring oscillatory excursions along repelling structures in phase space [Burke et al (2012), J. Math. Neurosci. 2, 3]; in particular, our work extends this study to systems featuring two slow variables and symmetry and goes significantly beyond the analysis presented in [Best et al (2005), SIADS 4, 1107-1139]. This is joint work with Kerry-Lyn Roberts (University of Sydney) and Jonathan Rubin (University of Pittsburgh). Host: Tom Chou, Ph.D.
Flyer Martin_Wechselberger_20141002.pdf

Proximal Distance Algorithms

Date Monday August 11, 2014 at 11:00 AM
Location 23-105 Center for the Health Sciences (CHS)
Speaker Ken Lange, Ph.D., Departments of Biomathematics and Human Genetics, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract The proximal distance principle is a new device for solving constrained optimization problems. The principle combines Clarke’s exact penalty method with distance majorization to create versatile algorithms effective even in discrete optimization. Proximal distance algorithms are highly modular and reduce to set projections and proximal mappings, both very well-understood techniques in optimization. Neither the objective function nor the constraint set need be convex. Initial results on linear programming, binary piecewise-linear programming, L0 regression, matrix completion, and inverse sparse covariance matrix estimation are very promising. Proximal distance algorithms are poised to play a major role in the high-dimensional optimization problems encountered in data mining, computational statistics, and bioinformatics.
Flyer Ken_Lange_20140811.pdf

From Microbiology to Microcontrollers: Effective collective search strategies in T cells and robotic swarms

Date Thursday May 08, 2014 at 4:00 PM
Location 23-105 Center for the Health Sciences
Speaker Melanie Moses, Ph.D., Associate Professor, Department of Computer Science, University of New Mexico
Sponsoring Dept UCLA Biomathematics
Abstract In order to trigger an adaptive immune response, T cells move through lymph nodes searching for dendritic cells that carry antigens indicative of infection. We observe that the distribution of step-sizes taken by T cells are heavy-tailed, Levy-like distributions that are characterized by many small steps and rare large steps. Levy walks have been shown to be efficient movement patterns for animals searching for prey. The heavy tailed distribution of step sizes in T cells are a collective property of millions of cells, even though individual T cells do not necessarily move using Levy walks. Our simulations show that T cell movements generate effective collective search strategies that dramatically improve the encounter rate with dendritic cells. It is not yet known whether the movement we observe is intrinsic to the T cells or is generated adaptively in response to extrinsic factors in the lymph node environment. We implement T cell movement patterns as a search strategy in a swarm of autonomous robots that collectively search for targets in their environment. We compare T cell movement to other collective search strategies and gain insights into how T cell interactions with features of the lymph node might increase encounter rates between T cells and dendritic cells.
Flyer Melanie_Moses_20140508.pdf

The Dynamics of Fibrin Gel Formation

Date Thursday March 27, 2014 at 4:00 PM
Location 13-105 Center for the Health Sciences (CHS)
Speaker James P. Keener, Ph.D., Distinguished Professor of Mathematics, Adjunct Professor of Bioengineering, University of Utah
Sponsoring Dept UCLA Biomathematics
Abstract Biogels are complex polymeric networks whose proper function is important to many physiological processes. For example, the proper function of mucus gel is important for airway clearance, reproduction, digestion, gastric protection, and disease protection and its failure is involved in cystic fibrosis, gastric ulcers, and reproductive dysfunction. Fibrin clots are crucial for prevention of bleeding after injury but inappropriate formation of clots is implicated in hearts attacks and strokes.
There are three phases of biogel dynamics that are important to their biological function. These are their formation (i.e., blood clotting), degradation (clot dissolution), and swelling/deswelling kinetics (during mucin secretion/exocytosis, for example).
The purpose of this talk is to describe recent advances in the study of the dynamics of fibrin clot formation. In particular, I will derive and discuss features of a new partial differential equation model that describes the growth of fibrin clots as a polymerization/gelation reaction. The solution of this PDE model gives insight into the branching structure of clots that are formed under various physiological conditions.
Host: Tom Chou, Ph.D.
To receive e-mail seminar notices, contact David Tomita (
Flyer Keener_seminar_20140327_.pdf

Quantitative Neurologic and

Date Wednesday March 19, 2014 at 1:00 PM
Location 24-132 Center for the Health Sciences, Morton Conference Room
Speaker Moses Wilks, Doctoral Graduate Student, Department of Biomathematics
Sponsoring Dept UCLA Biomathematics
Abstract Positron Emission Tomography (PET) is an inherently quantitative tool for measuring in vivo biological phenomena. However, there are still many barriers, both practical and structural, to robust quantifi cation of data in clinical and pre-clinical settings.
First, I present methods for improving quantifi cation of neurologic PET in Alzheimer’s disease imaging. Due to the variability in patient anatomy and disease state, it is diffi cult to accurately compare homologous structures between subjects. Here we examine methods of image normalization and automatic image analysis that allow for greatly reduced variance in data measurement. We show that through these methods, both the diagnostic and prognostic utility of the data can be greatly improved.
Additionally, we address the structural barriers to quantifi cation in oncologic PET in radio-labeled custom antibodies. These large, high-affi nity, tracers have been shown, both in silico and in vivo, to display high degrees of heterogeneous binding in target tissues. Due to this phenomenon, classical ODE models of tracer kinetics are no longer valid. We develop and test a new set of non-linear PDE models to accurately represent tracer activity in vivo. We show that the use of classical ODE models will result in high levels of parameter estimate bias, and the new PDE models can accurately fi t both in silico and in vivo data with the inclusion of Bayesian priors.
Doctoral Committee: Henry Huang, D.Sc. (Chair), Jorge R. Barrio, Ph.D., Elliot M. Landaw, M.D., Ph.D., Kenneth Lange, Ph.D., Anna M. Wu, Ph.D.
Flyer Moses_Wilks_defense_seminar.pdf

Phenotypic Bayesian phylodynamics: hierarchical graph models, antigenic clustering and latent liabilities

Date Wednesday February 05, 2014 at 9:30 AM
Location 5229 Life Sciences Building
Speaker Gabriela Cybis, Doctoral Graduate Student, Department of Biomathematics, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract Combining models for phenotypic and molecular evolution can lead to powerful inference tools. Under the fl exible framework of Bayesian phylogenetics, I develop statistical methods to address phylodynamic problems in this intersection. First, I present a hierarchical phylogeographic method that combines information across multiple datasets to draw inference on a common geographical spread process. Each dataset represents a parallel realization of this geographic process on a different group of taxa, and the method shares information between these realizations through a hierarchical graph structure. Additionally, I develop a multivariate latent liability model for assessing phenotypic correlation among sets of traits, while controlling for shared evolutionary history. This method can effi ciently estimate correlations between multiple continuous traits, binary traits and discrete traits with many ordered or unordered outcomes. Finally, I present a method that uses phylogenetic information to study the evolution of antigenic clusters in infl uenza. The method builds an antigenic cartography map informed by the assignment of each infl uenza strain to one of the antigenic clusters.
Flyer Cybis_defense_seminar.pdf

Some Problems and Opportunities in Biomedical Informatics

Date Thursday December 05, 2013 at 4:00 PM
Location 13-105 Center for the Health Sciences (CHS)
Speaker Douglas Bell, M.D., Ph.D., Associate Professor, Department of Medicine, Division of General Internal Medicine and Health Services Research, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract ABSTRACT: Federal policy is fostering a national wave of electronic health record (EHR) adoption, which opens a new source of data on patients’ illnesses, treatments and outcomes. For example, researchers are now using EHR data from large populations to conduct phenome-wide association studies (PheWAS) that seek to discover the full range of disease entities or outcomes associated with a particular gene or condition. However, uses of EHR data are limited by a lack of working semantic standards for many domains as well as poor application of the semantic standards that exist. I will review the history of work on controlled vocabularies and ontologies that seek to represent health-related concepts in a computer-interpretable form. EHR systems also create opportunities to influence health care practice patterns by delivering information to providers at the time of decision-making. However, in practice, such alerts often have weak effects, in part because they are triggered inaccurately and also because providers misunderstand their importance. Better models of provider knowledge and their actions under time constraints might substantially advance the efficacy of clinical decision support. Forthcoming changes in health care financing, such as accountable care organizations, may help to align incentives toward finding solutions to these problems.
Flyer Douglas_Bell_20131205.pdf

Fast Spatial Ancestry Estimation via Flexible Allele Frequency Surfaces

Date Monday November 25, 2013 at 2:00 PM
Location Gonda Center 5303
Speaker John Michael Ranola, Doctoral Graduate Student, Department of Biomathematics, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract Unique modeling and computational challenges arise in locating the geographic origin of individuals based on their genetic backgrounds. SNPs vary widely in informativeness, allele frequencies change nonlinearly with geography, and reliable localization requires evidence to be integrated across a multitude of SNPs. These problems become even more acute for individuals of mixed ancestry and present substantial computational challenges that have been addressed with approximate models. We attack these problems by borrowing ideas from image processing and optimization theory. Our model discretizes the region of interest into pixels and operates SNP by SNP. We estimate allele frequencies across the landscape by maximizing a product of binomial likelihoods penalized by nearest neighbor interactions. Penalization smooths allele frequency estimates and promotes estimation at pixels with no data. Maximization is accomplished by an MM (minorize-maximize) algorithm. Once allele frequency surfaces are available, one can apply Bayes rule to compute the posterior probability that each location is the origin of a given person. Placement of admixed individuals on the landscape is more complicated and requires estimation of the fractional contribution of each pixel to a person’s genome. This estimation problem also succumbs to a penalized MM algorithm. On the POPRES data, the current model gives better localization for both unmixed and admixed individuals than existing methods despite using just a small fraction of the available SNPs. Computing times are comparable to the best competing software.
Doctoral Committee: Kenneth Lange, Ph.D. (chair), Janet Sinsheimer, Ph.D., Marc Suchard, M.D., Ph.D., Steve Horvath, Ph.D., Sc.D.

To receive e-mail seminar notices, contact David Tomita (
Flyer John_Michael_Ranola_defense_seminar.pdf

Theory of Hair Bundle Motion; New Insights from the Simplest Models

Date Thursday November 14, 2013 at 4:00 PM
Location 13-105 Center for the Health Sciences (CHS)
Speaker Seung Ji, Doctoral Graduate Student, Department of Physics & Astronomy, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract Normal Form of Hopf bifurcaiton equation(NFE) has been used to describe a Hair bundle motion. In this talk, I will show that the NFE can also describe the mode-locking mechanism of hair bundle and NFE with noise exhibits stochastic resonance. Also, I propose that Kuramoto model can describe dynamics of the coupled hair bundles and stochastic resonance can play essential role in the detection of sound.
Host: Van Savage, Ph.D.
To receive e-mail seminar notices, contact David Tomita (
Flyer Seung_Ji_20131114.pdf

Integration of high throughput miRNA and mRNA data through weighted gene co-expression network analysis

Date Thursday March 07, 2013 at 4:00 PM
Location 13-105 Center for the Health Sciences (CHS)
Speaker David Elashoff, Ph.D., Professor of Medicine and Biostatistics, Director of the Department of Medicine Statistics Core, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract miRNA regulates mRNA levels through base-pairing, by inducing transcript degradation or by inhibiting translation. Many high throughput biological experiments simultaneously assess global miRNA and mRNA profiles and seek to find coordinate expression modifications in both types of data. There are many computational algorithms to integrate miRNAs with their putative gene targets. We developed a method using weighted gene co-expression network analysis to identify gene targets for each miRNA. This method is illustrated with an example in a renal carcinoma dataset from patients with matched normal and tumour samples. We also find that by using WGCNA to define highly correlated genes into a number of modules greatly alleviates the multiple testing problems that plague standard gene-centric methods and it provides a novel integrative view of miRNAs and their putative genes.
Flyer speaker_David_Elashoff_20130307.pdf

Large-Deviations Theory for Living Systems

Date Thursday December 06, 2012 at 4:00 PM
Location 13-105 Center for the Health Sciences (CHS)
Speaker D. Eric Smith, Ph.D., External Professor, Santa Fe Institute, Santa Fe, New Mexico
Sponsoring Dept UCLA Biomathematics
Abstract The separation of timescales, in systems whose state spaces are highly structured, is a major source of hierarchy, complexity, and robustness in living systems. Fluctuations or excursions that are rare on the timescale of microscopic processes may be the key functional events at the next higher scale of aggregation, whose characteristic timescale may be exponentially (in the relative scales of aggregation) longer than the microscopic timescale. Examples include rate-limiting steps in catalysis, checkpoints in signaling or cell division, and perhaps improbable (but not infinitely improbable!) steps in the emergence of life. Quantitative estimation of the rate and structure of rare events in biology requires new methods beyond those familiar from classical statistical mechanics, because living systems are out of equilibrium, because numbers of degrees of freedom are often mesoscopic, and because their heterogeneity precludes many simplifications from high degrees of symmetry. In this talk I will first introduce principles and methods from dynamical large-deviations theory, which apply to such problems and are also at the interface with recent active work in non-equilibrium statistical mechanics. I will then review a few biological applications, some of which have been solved, and others (particularly in the origin of life) which are suggestive but are still looking for an adequate empirical and mathematical framing of the right questions.
Flyer smith_eric_20121206.pdf

Highly Accurate Consensus Sequence and Genome Assembly using PacBio® RS Sequence Data

Date Thursday November 29, 2012 at 4:00 PM
Location 13-105 Center for the Health Sciences (CHS)
Speaker David Alexander, Ph.D., Senior Algorithms Engineer, Pacific Biosciences, Menlo Park, CA
Sponsoring Dept UCLA Biomathematics
Abstract Pacific Biosciences’ single molecule real-time (SMRT®) DNA sequencing platform provides extremely long read lengths and a rich profile of DNA template modifications, enabling advances in de novo genome assembly and epigenetic studies. In this talk I will discuss some of the basic mathematical models behind genome assembly, and then give an overview of our hierarchical genome assembly procedure (HGAP). After a brief overview of the properties of PacBio sequence data, I will describe our algorithm, Quiver, for highly accurate consensus sequence determination, which can generate finished assemblies with accuracies exceeding 99.999%.
Flyer david_alexander_20121129.pdf

Modeling Tumor Growth in an Evolving Organ: a Diffuse Domain Approach

Date Thursday November 15, 2012 at 4:00 PM
Location 13-105 Center for the Health Sciences (CHS)
Speaker Yao-Li Chuang, Ph.D., Postdoctoral Fellow, Department of Pathology, Computational Biology Program, University of New Mexico
Sponsoring Dept UCLA Biomathematics
Abstract Tumor growth at the macroscopic (tissue) scale is often characterized by emerging spatiotemporal patterns. The induced heterogeneity can have further impact on the interaction between tumor cells and their microenvironment, causing the tumor to develop infiltration behaviors. Many partial-differential-equation (PDE) models have been proposed to study tumor progression, describing the interaction between the tumor cells and the substrate concentrations as diffusion-reaction processes. In “Three-dimensional multispecies nonlinear tumor growth - I Model and numerical method” [Wise et al., J. Theor. Biol. 253, pp.524-543 (2008)], we used Cahn-Hilliard equations to model tumor progression, accounting for the differentiated cell adhesion that causes phase separation among cells. The model successfully connected the development of tumor infiltration patterns to the microenvironmental stress. Like many other PDE tumor growth models, the Cahn-Hilliard model was formulated on an unbounded domain. In living tissues, however, tumor growth is often bounded by the tissues within an organ, which may in turn evolve due to the pressure of tumor growth. To describe the growth of such tumors, we use a recently developed diffusion domain approach, which allows us to adapt a PDE model defined on an unbounded domain to an evolving confinement. In this talk, I will first describe our tumor growth model and review our recent applications and findings of the model. Then I will introduce the diffuse domain approach and show how we adapt the tumor growth model using lymphoma progression in a lymph node and breast tumors in a mammary duct as examples.
Flyer yao-li_chuang_20121115.pdf

Automated measurements from images and video to enable better modeling: Vascular geometry, consumer-resource interactions, and bacterial colony sizes

Date Thursday November 01, 2012 at 4:00 PM
Location 13-105 Center for the Health Sciences
Speaker Van Savage, Ph.D. and Pamela Yeh, Ph.D., V. Savage, Ph.D., Asst. Prof., UCLA Dept. of Biomathematics; P. Yeh, Ph.D., Adj. Asst. Prof., UCLA Dept. of Env. Health Sciences
Sponsoring Dept UCLA Biomathematics
Abstract The promise of using video and images to collect biological data has been a major research focus for decades. Enough tools and algorithms now exist that it is possible for researchers to straightforwardly extract data from images and video for physiological and ecological systems. These approaches enable high-throughput methods that often yield higher-quality data and substantially more data than previous efforts. Moreover, these approaches can also enable the collection of new types of data. We will briefly discuss three examples of systems from which we have gathered new data using these methods. First, we will describe new software for automatically measuring vessel dimensions and geometry from three-dimensional angiographic (e.g., CT and MRI) images. This software leads to much faster collection of larger amounts of data. Because these angiographic images are non-invasive, they do not distort or destroy the vasculature and thus lead to more exact measurements. Second, we show how video-tracking software can be used to track wingless fruit flies that are being hunted by wolf spiders. This software leads to measurements of components of consumer-resource interactions that have rarely or never been measured before and with much higher resolution. Finally, and as our primary focus for this talk, we discuss new software to identify individual bacterial colonies grown in agar plates, measure their sizes, and construct size distributions. With these more accurate and larger volumes of data, we can analyze how the phenotype of colony size responds to a range of concentrations of antibiotics. Rather than just focusing on the mean, we now have sufficient data to investigate the variance, coefficient of variation, and other measures of the distribution, in the hopes of revealing new insights about bacterial diversity and the evolution of resistance.
Flyer Savage_Yeh_20121101.pdf

Evolution of Aging, Mortality and Immortality in Bacteria

Date Thursday October 25, 2012 at 4:00 PM
Location 13-105 Center for the Health Sciences (CHS)
Speaker Lin Chao, Ph.D., Professor, Section of Ecology, Behavior and Evolution, Division of Biological Sciences, UC San Diego
Sponsoring Dept UCLA Biomathematics
Abstract Deleterious mutations appearing in a population increase in frequency until stopped by natural selection. The ensuing equilibrium creates a stable frequency of deleterious mutations or the mutational load. Here I develop the comparable concept of a damage load, which is caused by harmful non-heritable changes to the phenotype. A damage load also ensues when the increase of damage is opposed by selection. The presence of a damage load favors the evolution of asymmetrical transmission of damage by a mother to her daughters. The asymmetry is beneficial because it increases fitness variance, but it also leads to aging or senescence. A mathematical model based on microbes reveals that a cell lineage dividing symmetrically is immortal if lifetime damage rates do not exceed a threshold. The evolution of asymmetry allows the lineage to persist above the threshold, but the lineage becomes mortal. In microbes with low genomic mutation rates, it is likely that the damage load is much greater than the mutational load. In metazoans with higher genomic mutation rates, the damage and the mutational load could be of the same magnitude. A fit of the model to experimental data shows that Escherichia coli cells experience a damage rate that is below the threshold and are immortal under the conditions examined. The model estimates the asymmetry level of E. coli to be low but sufficient for persisting at higher damage rates. The model also predicts that increasing asymmetry results in diminishing fitness returns, which may explain why the bacterium has not evolved higher asymmetry.
Flyer lin_chao_20121025.pdf

A Mathematical Model of Intimal Thickening: an Application to Atherosclerosis

Date Thursday October 18, 2012 at 4:00 PM
Location 13-105 Center for the Health Sciences (CHS)
Speaker Pak-Wing Fok, Ph.D., Assistant Professor, Department of Mathematical Sciences, University of Delaware
Sponsoring Dept UCLA Biomathematics
Abstract Atherosclerosis is an inflammatory disease of the artery characterized by an expansion of the intimal region. Intimal thickening is usually attributed to the migration of smooth muscle cells (SMCs) from the surrounding media and proliferation of SMCs already present in the intima. Intimal expansion can give rise to dangerous events such as stenosis (leading to stroke) or plaque rupture (leading to myocardial infarction). We propose and study a mathematical model of intimal thickening, posed as a free boundary problem. Intimal thickening is driven by damage to the endothelium, resulting in the release of cytokines and migration of SMCs. By coupling a boundary value problem for cytokine concentration to an evolution law for the intimal area, we reduce the problem to a single nonlinear differential equation for the luminal radius. We analyze the steady states, perform a bifurcation analysis and compare model solutions to data from rabbits whose iliac arteries are subject to a balloon pullback injury. In order to obtain a favorable fit, we find that migrating SMCs must enter the intima very slowly compared to cells in dermal wounds. This cell behavior is indicative of a weak inflammatory response which is consistent with atherosclerosis being a chronic inflammatory disease.
Flyer fok_pak-wing_20121018.pdf

Networks of Neurons Create Complex Dynamics: Statistical Physics and a Simple Model for the Control of Breathing

Date Thursday October 11, 2012 at 4:00 PM
Location 13-105 Center for the Health Sciences (CHS)
Speaker Alex Levine, Ph.D., Professor, Departments of Chemistry & Biochemistry and Physics & Astronomy, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract ABSTRACT: “Cogito ergo sum.” There is a long history of physical scientists thinking about thinking, going back at least as far as Rene Descrates’ famous pronouncement. Much more recently, a combination of neuroscientists and physicists realized that it is possible to explore the dynamics of interacting neurons using ideas borrowed from nonlinear dynamical systems and statistical mechanics. In particular the nervous systems contains many reasonably small collections of neurons that collectively generate well-defined pattern of electrical activity, which continue even when those collections of cells are removed from the animal. These functional groups of neurons are now termed central pattern generators. While understanding such restricted systems does not necessarily elucidate such sublime questions as those regarding the nature of consciousness, these studies do provide an intriguing example of a novel application of statistical mechanics to biology. They also admit quantitative comparisons to experiment! In this talk I present a minimal model for one such central pattern generator based on the interaction of nonlinear dynamical systems interacting on a quenched random network. No neuroscience background will be assumed and, fortunately, very little will be required for exploring how a simple model of coupled excitatory neurons can produce collective and metronomic bursts of activity that controls the breathing rhythm in mammals. I will focus on how topological properties of the random network of neuronal connections controls the collective dynamical phase diagram of the system. I will conclude with some new extensions of this work to the building of similarly simple models of the global and rhythmic dynamics of the neocortex, the seat of consciousness and the paragon of complexity that produced “Cogito ergo sum.”
Flyer alex_levine_20121011c.pdf

Dynamic Mutation Selection Balance as an Evolutionary Attractor

Date Thursday October 04, 2012 at 4:00 PM
Location 13-105 Center for the Health Sciences (CHS)
Speaker Sidhartha Goyal, Ph.D., Postdoctoral Scholar, The Kavli Institute for Theoretical Physics,UC Santa Barbara
Sponsoring Dept UCLA Biomathematics
Abstract ABSTRACT: The vast majority of mutations are deleterious, and are eliminated by purifying selection. Yet in finite asexual populations, purifying selection cannot completely prevent the accumulation of deleterious mutations due to Muller’s ratchet and can lead to a rapid degradation of population fitness. Evidently, the long term stability of an asexual population requires an in flux of beneficial mutations and any stable evolutionary state of a population in a static environment must involve a dynamic mutation-selection balance, where accumulation of deleterious mutations is on average offset by the in flux of beneficial mutations. We find that a surprisingly low fraction of beneficial mutations suffices to achieve stability, even in small populations in the face of high mutation rates and weak selection. This may explain the maintenance of mitochondria and other asexual genomes.
Flyer sidhartha_goyal_20121004.pdf

Modeling cancer stem cell state transitions and extinction

Date Thursday June 07, 2012 at 4:00 PM
Location 23-105 Center for the Health Sciences
Speaker Mary Sehl, M.D., Ph.D., Clinical Instructor, Department of Medicine, Department of Medicine, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract Targeted therapy dramatically improves survival in breast cancer patients whose tumor overexpresses HER2. A subpopulation of cells in human breast tumors has been identified with characteristics of cancer stem cells. Microenvironmental signaling guiding epithelial-to-mesenchymal transition (EMT) and the reverse process (MET) is thought to play a role in the plasticity of these breast cancer stem-like cells (BCSCs). BCSCs rely on HER2 signaling for self-renewal, suggesting that HER2-targeted therapy targets BCSCs even when the bulk of the tumor does not overexpress HER2. In order to guide clinical trials examining HER2-targeted therapy in the adjuvant setting, we propose a mathematical model to examine BCSC population dynamics and predict optimal duration of therapy. Varying the susceptibility of BCSCs to HER2-targeted therapy, we quantify the average time to extinction of BCSCs. We expand our model using stochastic simulation to include the partially differentiated tumor cells (TCs) that represent bulk tumor population and examine effects of plasticity on required duration of therapy. Lower susceptibility of BCSCs and increased rates of dedifferentiation entail longer :extinction times, indicating a need for prolonged administration of HER2-targeted therapy. We predict that even when therapy does not appreciably reduce tumor size in the advanced cancer setting, it will eventually eradicate the tumor in the adjuvant setting as long as there is at least a modest effect on BCSCs. We anticipate that our results will inform clinical trials of targeted therapies in planning the duration of therapy needed to eradicate BCSCs. Our predictions also address safety, as longer duration of therapy entails a greater potential impact on normal stem cells that may also be susceptible to stem cell-targeted therapies. We expand our model to study the complex regulatory feedback involved in stem cell niche regulation and EMT/MET transitions. Host: Ph.D.: Elliot Landaw, M.D., Ph.D. To receive e-mail seminar notices, contact David Tomita (
Flyer sehl_mary_20120607.pdf

Genome-Based HIV-1 Incidence Assay with High Sensitivity and Specificity

Date Thursday May 17, 2012 at 4:00 PM
Location 23-105 Center for the Health Sciences (CHS)
Speaker Ha Youn Lee, Ph.D., Associate Professor, Dept. of Molecular Microbiology, University of Southern California
Sponsoring Dept UCLA Biomathematics
Abstract ABSTRACT: To assess how many people have been recently infected in a given area is an important task in HIV-1/AIDS prevention. Accurately classifying recent or incident infections (e.g. within around the first year since transmission) from chronic infections enables one to track the epidemics, evaluate the impact of antiretroviral treatment, and assess the efficacy of HIV-1 prevention trials including vaccination, microbicides, and other types of interventions. Conventional serological testing is found to have a number of critical limitations which result in notable inaccuracy. In this talk, we turn to utilizing recent advances in understanding early HIV-1 infections and demonstrate that information derived from a set of HIV-1 envelope gene sequences obtained from a single blood sample can accurately distinguish incident infections from chronic ones. By analyzing previously published 5596 full envelope HIV-1 genes from 182 incident and 43 chronic subjects, we find that every incident case displays a robust signature, the presence of closely related strains, regardless of either single-variant or multi-variant transmission. We demonstrate that the sequence similarity used as a biomarker has high specificity and sensitivity over 95% and is not sensitive to viral and host specific factors including the clade of the viral strain, viral load, and the length and location of sequences in the HIV-1 envelope gene. The potency and accuracy of our sequencing-based HIV incidence assay is unprecedented and the assay holds great promise as a means of assessing the level of HIV-1 incidence from a single blood draw in cross-sectional blood surveys.
Flyer lee_ha_youn_20120517.pdf

CANCELLED - Network controls on diversity in human and ecological systems

Date Thursday April 05, 2012 at 4:00 PM
Location 23-105 Center for the Health Sciences (CHS)
Speaker Evan Economo, Ph.D., Assistant Professor, Dept. of Ecology & Evolutionary Biology, Michigan Society of Fellows, Univ. of Michigan
Sponsoring Dept UCLA Biomathematics
Abstract Both ecological and human systems are characterized by stunning patterns of diversity. Understanding biodiversity- the variation of genes, species, phenotypes and interactions within and across ecosystems, has long been a fascination of biologists. Humans have created worlds of diversity as well, in the foods we eat, the products we buy, and most of all in the ideas we think. Ecosystems are arranged spatially and connected in complex ways by dispersal of individual organisms, and humans communicate through interpersonal and technological networks. In this talk I explore the theoretical connections between these two types of systems by integrating network influence models in social science with coalescent theory in population genetics. I identify several general features of network structure that can amplify or depress diversity. In particular, the asymmetries of connections among nodes emerges as more important variation in the strength and topology of connectivity. With these theoretical perspectives, I analyze a parameterized coral reef network across the Pacific and show that certain nodes act as structural depressors due to their position in the network. Finally, I discuss how recent revolutions in information transmission may be changing the diversity, performance, and collective wisdom of human societies. Network controls on diversity in human and ecological systems.
Flyer economo_evan_20120405_cancelled.pdf

Stochastic nucleation and growth

Date Thursday February 23, 2012 at 4:00 PM
Location 23-105 Center for the Health Sciences (CHS)
Speaker Maria Rita D'Orsogna, Ph.D., Assistant Professor, Mathematics Department, CSU Northridge
Sponsoring Dept UCLA Biomathematics
Abstract The binding of individual components to form composite structures is a ubiquitous phenomenon within the sciences. Within heterogeneous nucleation, particles may be attracted to an initial exogenous site: the formation of droplets, aerosols and crystals usually begins around impurities or boundaries. Homogeneous nucleation on the other hand describes identical particles spontaneously clustering upon contact. Given their ubiquity in physics, chemistry and material sciences, nucleation and growth have been extensively studied in the past decades, often assuming infinitely large numbers of building blocks and unbounded cluster sizes. These assumptions also led to the use of mass-action, mean field descriptions such as the well known Becker Doering equations. In cellular biology, however, nucleation events often take place in confi ned spaces, with a fi nite number of components, so that discrete and stochastic effects must be taken into account. In this talk we examine finite sized homogeneous nucleation by considering a fully stochastic master equation, solved via Monte-Carlo simulations and via analytical insight. We find striking differences between the mean cluster sizes obtained from our discrete, stochastic treatment and those predicted by mean field treatments. We also consider heterogeneous nucleation stochastic treatments, first passage time results and possible applications to prion unfolding and clustering dynamics.
Flyer maria_dorsogna_20120223.pdf


Date Thursday December 01, 2011 at 4:00 PM
Location 23-105 Center for the Health Sciences (CHS)
Speaker Eric Sobel, Ph.D., Adjunct Professor, Department of Human Genetics, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract To be Announced

Hemodynamic Optimization Parameter Evaluation for Cardiac Resynchronization Therapy

Date Thursday November 17, 2011 at 4:00 PM
Location 23-105 Center for the Health Sciences (CHS)
Speaker Gene A. Bornzin, Ph.D., Vice President, Fellow, Research, St. Jude Medical, Inc., Sylmar, California
Sponsoring Dept UCLA Biomathematics
Abstract Cardiac Resynchronization Therapy (CRT) improves cardiac performance in advanced heart failure patients. Patients indicated for CRT have low ejection fraction and a wide electrocardiographic QRS. These patients have a mechanically dyssynchronous contraction of the ventricles that can be improved by providing cardiac stimulation of both the right and left ventricles using specialized cardiac pacemakers. CRT pacing therapy reduces cardiovascular related hospitalizations and improves survival while decreasing symptoms and improves exercise tolerance. About 200,000 patients benefit from CRT annually. This presentation will cover CRT therapy delivery including the basic pacemaker and defibrillation devices, implantation, and some technological challenges remaining. One challenging clinical objective is not just to provide therapy, but to optimize therapeutic benefit. Currently in select patients, pacemaker settings are adjusted to optimize hemodynamic performance using echo cardiography. This process takes place in the clinic and is extremely time consuming and expensive. Ideally, CRT pacemaker settings could be quickly and easily optimized for every patient during a routine office visit. Ultimately, in the not too distant future, the implanted CRT devices will actually measure cardiac performance with sensing systems. Feedback from the sensing system would then be used to adjust stimulation parameters to optimize cardiac performance. Completed research into methods that could be used to measure and optimize cardiac performance will be presented along with a discussion of ongoing clinical research.
Flyer Bornzin_Gene_2011117.pdf

A Parthian shot at neutrality: revisiting the neutrality assumption for tropical tree species

Date Thursday November 10, 2011 at 4:00 PM
Location 23-105 Center for the Health Sciences (CHS)
Speaker Jerome Chave, Ph.D, Senior Researcher (Director) at National Center for Scientific Research, France
Sponsoring Dept UCLA Biomathematics
Abstract ABSTRACT: Hubbell’s neutral theory of biodiversity challenges the classical niche-based view of ecological communities, where species attributes and environmental conditions jointly determine community composition. Functional equivalence among species, as assumed by neutral ecological theory, has been recurrently falsified, yet many patterns of tropical tree communities appear consistent with neutral predictions. This may mean that neutral theory is a good first-approximation theory or that species abundance data sets contain too little information to reject neutrality. Here we present a simple test of neutrality based on species abundance distributions in ecological communities. Based on this test, we show that deviations from neutrality are more frequent than previously thought in tropical forest trees, especially at small spatial scales. We then develop a nonneutral model that generalizes Hubbell?s dispersal-limited neutral model in a simple way by including one additional parameter of frequency dependence. We also develop a statistical method to infer the parameters of this model from empirical data by approximate Bayesian computation. In more than half of the permanent tree plots, we show that our new model fits the data better than does the neutral model. Finally, we discuss whether observed deviations from neutrality may be interpreted as the signature of environmental filtering on tropical tree species abundance distributions. This study is mostly based on F Jabot and J Chave, 2011, Analyzing Tropical Forest Tree Species Abundance Distributions Using a Nonneutral Model and through Approximate Bayesian Inference American Naturalist. with some background information on the biological question for the broader audience of a biomathematics seminar.
Flyer jerome_chave_20111110.pdf

Using Genomes to Track the Evolution of Life on Earth and Beyond

Date Thursday November 03, 2011 at 3:00 PM
Location Schoenberg Hall, Schoenberg Music Building
Speaker James Lake, Ph.D., Faculty Research Lectureship, 11/3/11
Sponsoring Dept 111th UCLA Faculty Research Lectureship
Description We are referring guests to the 111th UCLA Faculty Research Lecture being given by Prof. James Lake, Distinguished Professor of Molecular, Cell and Developmental Biology and Human Genetics. See above Sponsoring Department link.

The biofluiddynamics of fungal spore and genome dispersal

Date Thursday October 20, 2011 at 4:00 PM
Location 23-105 Center for the Health Sciences (CHS)
Speaker Marcus Roper, Ph.D., Professor, Department of Mathematics, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract ABSTRACT: Fungi are the most diverse of all eukaryotic organisms and enjoy extraordinary ecological success as decomposers, pathogens and mutualists. I will discuss how solving hard physical problems of dispersing (i) spores and (ii) genomes may be a central part of their success in so many niches: #1. The forcibly launched spores of ascomycete fungi must eject through a boundary layer of nearly still air in order to reach dispersive air flows. Because of their microscopic size singly ejected spores are almost immediately brought to rest by fluid drag. However, by coordinating the ejection of thousands or hundreds of thousands of spores, fungi such as the devastating plant pathogen Sclerotinia sclerotiorum, are able to create a flow of air that carries spores across the boundary layer and around any intervening obstacles. #2. A growing filamentous fungi may harbor a diverse population of nuclei. Increasing evidence shows that this internal genetic flexibility is a motor for diversification, virulence, and the ability of fungi to utilize nutritionally complex substrates like plant cell walls. I’ll show that to maintain stable populations of different nuclei near the growing tips, ascomycete fungi must create internal flows over the entire of the colony.
Flyer roper_marcus_20111020.pdf

Spatially-explicit ecological dynamics in streams and rivers

Date Thursday October 13, 2011 at 4:00 PM
Location 23-105 Center for the Health Sciences (CHS)
Speaker Kurt Anderson, Ph.D., Assistant Professor, Department of Biology, UC Riverside
Sponsoring Dept UCLA Biomathematics
Abstract Many organisms disperse in media possessing a net unidirectional flow. The systems these organisms inhabit, exemplified by streams and rivers, are also characterized by a high degree of multi-scale spatial and temporal environmental variability. Most conceptual frameworks describing ecological organization in streams and rivers prominently feature both upstream-downstream linkages and variability that occurs across spatial and temporal scales. I will discuss modeling studies where I have explored how the spatial distribution of organisms results from multi-scale spatial variability in systems with directionally-biased dispersal. I will begin by discussing spatial scales that characterize population responses near equilibrium. Then, I will discuss transient and non-equilibrium dynamics using metrics that are independent of initial conditions – resilience, reactivity, and the amplification envelope – and relate them to the spatial scale of the population perturbation. Current work aims to extend previous themes to branching river networks and the surrounding landscape. I will conclude with implications for conservation of instream populations.
Flyer anderson_kurt_20111013.pdf

Scaling up the effects of physiological constraints from individuals to communities

Date Thursday October 06, 2011 at 4:00 PM
Location 23-105 Center for the Heatlh Sciences (CHS)
Speaker Samraat Pawar, Ph.D., Postdoctoral Fellow, Department of Biomathematics, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract Biodiversity patterns are generated by processes acting at multiple levels of biological organization, ranging from individual organisms to whole ecosystems consisting of multiple, interactions populations. Understanding the mapping between these levels is crucial for the development of a general theory of the generation and maintenance of biodiversity. I present a quantitative framework for predicting how individual-level physiological constraints in nature drive the dynamics and structure of multiple species communities. This theory accurately predicts the effects of organismal body size, habitat spatial dimensionality and environmental temperature on consumer-resource interactions, and scales up these interactions to the species interaction networks that drive dynamics of whole communities. Furthermore, it makes predictions that explain a number of observed features of existing biodiversity patterns, and can form a foundation for better predicting future changes in these patterns due to natural and anthropogenic changes in the environment.
Flyer pawar_samraat_20111006.pdf

Using model-based methods to quantify exon-level gene expression from RNA-seq data

Date Thursday September 29, 2011 at 4:00 PM
Location 23-105 Center for the Health Sciences (CHS)
Speaker Zhaohui Steve Qin, Ph.D, Associate Professor, Department of Biostatistics and Bioinformatics, Rollins School of Public Health, Emory University.
Sponsoring Dept UCLA Biomathematics
Abstract RNA sequencing (RNA-seq) is a powerful new technology for mapping and quantifying transcriptome using ultra high-throughput next generation sequencing technologies. Using deep sequencing, gene expression levels of all transcripts including novel ones can be quantified digitally. Although extremely promising, the massive amounts of data generated by RNA-seq, substantial biases, and uncertainty in short read alignment pose daunting challenges for data analysis. In particular, large base-specific variations and between-base correlations make simple approaches, such as those that use averaging to normalize RNA-seq data and quantify gene expressions, ineffective. In this study, we propose a model-based method to characterize base-level read coverage within each exon. The underlying expression level is included as a key parameter in this model. Since our method is capable of capturing local genomic features that affect read coverage profile throughout the exon, we are able to obtain improved quantification of the true underlying expression levels.
Flyer zhaohui_steve_qin_20110929_updated.pdf

Geometry and scaling in the vascular system: Theory vs MRI

Date Thursday September 01, 2011 at 4:00 PM
Location AV-139 Center for the Health Sciences (CHS)
Speaker Mitchell Johnson, Graduate Student, Department of Biomathematics, UCLA
Sponsoring Dept UCLA Biomathematics
Abstract The geometry of vascular systems determines the cost of energy delivery, and thus constrains the growth and metabolic rate of individuals, cells and tumors. We develop an algorithm and software package for automatically quantifying the geometry of vascular trees from radiography images (MR, CT, etc.). We use this technique to compare measurements for human arterial networks to theoretical predictions and previous measurements in humans, pigs, and rats, in order to address basic questions in the optimality of vessel network structure and metabolic scaling theory. Our goal is to develop high-throughput angiography, which would make it practical to gather data at a faster pace, and with higher precision, than is practical with existing methods.
Flyer johnson_mitchell_seminar_20110901.pdf

Joint Analysis of Longitudinal Measurements and Competing Risks Failure Time Data”

Date Monday August 29, 2011 at 2:00 PM
Location 53-105 Center for the Health Sciences (CHS)
Speaker Ning Li, Ph.D., Samuel Oschin Comprehensive Cancer Institute, Cedars-Sinai Medical Center
Sponsoring Dept UCLA Biomathematics
Flyer ning_li_20110829.pdf

Temperature Dependency of Energy and Mass Fluxes in Dynamic Energy Budget Theory.

Date Thursday May 26, 2011 at 4:00 PM
Location 43-105 Center for the Health Sciences (CHS)
Speaker Erik Muller, Ph.D., Associate Researcher, Marine Science Institute, Department of Ecology, Evolution and Marine Biology , UC Santa Barbara
Sponsoring Dept UCLA Biomathematics
Abstract Dynamic Energy Budget (DEB) theory is a process based theory that describes the rates at which an organism acquires resources from the environment and subsequently utilizes the energy and nutrients therein for production and maintenance. The core model covers all life stages of heterotrophic organisms with just 3 state variables and 12 parameters. Despite its focus on processes, the theory currently describes the impact of temperature on the dynamics of energy acquisition and allocation in a purely descriptive manner. In order to improve realism in the representation of temperature effects in DEB theory, I am using formalism from complex network theory. Before presenting this work in detail, I will survey the implications of the theory for seemingly unrelated biological phenomena, such as body-size scaling relationships, symbiogenesis and toxic effects.
Flyer muller_erik_20110526.pdf

Statistical and Computational Methods for Ancestry Estimation and Variable Selection in Genome-Scale Datasets

Date Friday May 06, 2011 at 11:00 AM
Location 14-214U Center for the Health Sciences
Speaker David Alexander, UCLA Department of Biomathematics
Sponsoring Dept UCLA Biomathematics
Abstract ABSTRACT: As genotyping and sequencing technologies reach higher and higher throughput levels, genetic datasets are becoming ever larger, creating a growing need for highly efficient algorithms for routine analyses. Our work on the efficient individual ancestry estimation program ADMIXTURE has shown that easily-implemented and stable EM algorithms, widely believed to be a good choice for estimation in large datasets, can sometimes prove vastly inferior to more intricate coordinate- and block-relaxation approaches. Furthermore, our work on a novel quasi-Newton convergence acceleration procedure shows that the efficiency of existing iterative optimization algorithms can be greatly improved with no change to the statistical model and only minor implementation effort. For secondary analyses, computationally intensive methods are more tolerable. In this vein, we explore the application of some new developments in bootstrap aggregation and high-dimensional variable selection to genome-wide association, to see whether more complex models can be used to wring more information from previously studied association data.
Flyer alexander_david_seminar.pdf