Off Week Sessions

Off Week Session Summary

We have created this page to organize the off week session talks. If you are not scheduled, do not panic. There are still open slots. If you have not emailed Valerie and Nick, please do so by responding to the original email sent requesting abstracts. If you are scheduled, please check to make sure the information is correct and that the dates work for you. Also, if you are missing information such as titles or abstracts or last names, please please contact us ASAP, so that we can correct these matters.

Schedule

September 27, 2011

Name: Nishanth Ulhas Nair (15 minutes)

Research Area: Computational Epigenetics

Title: ChIPnorm: a statistical method for normalizing and identifying differential regions in histone modification ChIP-seq libraries

Abstract

With the advent of high-throughput technologies such as ChIP-seq, the study of histone modifications has become of great interest. A problem of particular interest is the identification of regions of the genome where different cell types from the same organism exhibit different patterns of histone enrichment. This problem turns out to be surprisingly difficult, even in simple pairwise comparisons, because of the significant level of noise in ChIP-seq data. In this paper we propose a two-stage statistical method, called ChIPnorm, to normalize ChIP-seq data, and to find differential regions in the genome, given two libraries of histone modifications of different cell types. We show that the ChIPnorm method removes most of the noise and bias in the data compared with other normalization methods. We correlate the histone marks with gene expression data and confirm that histone modification H3K27me3 and H3K4me3 acts as a repressor and an activator of genes respectively. We show that a sizable fraction of bivalent marks in ES cells for H3K27me3 and H3K4me3 moves into a K27-only state possibly reflecting permanent repression of these developmental genes in the differentiated cell.
Name: Simon Focart (30 minutes)

Research Area: Compressive Sensing, Approximation Theory

Title: An overview of Compressive Sensing

Abstract

I will briefly describe the mathematics of Compressive Sensing, where one aims at reconstructing sparse vectors from a seemingly incomplete set of linear measurements. The focus will be put on the reconstruction process rather than the measurement process. I will review the $\ell_1$-minimization, the orthogonal matching pursuit, and the hard thresholding pursuit algorithms.
Name: Catherine Grasso (15 minutes)

Research Area: Exome Sequencing

Title: Exome sequencing and integrative mutational profiling of lethal castrate resistant prostate cancer

Abstract

Characterization of the prostate cancer transcriptome and genome has identified chromosomal rearrangements and copy number gains/losses, including ETS gene fusions, PTEN loss and AR amplification, that drive prostate cancer development and progression to lethal metastatic castrate resistant prostate cancer (CRPC). As less is known about the role of simple mutations, here we sequenced the exomes of 21 lethal CRPCs (including three different foci from the same patient) and 3 localized prostate cancers, which demonstrated low overall mutation rates even in heavily treated CRPC (2.23/Mb), confirmed the monoclonal origin of lethal CRPC, identified likely deleterious somatic mutations in genes that interact with AR, and identified three known genes (AR, TP53, and ZFHX3) as significantly mutated in prostate cancer. Through integration with a compendium of related genomic data, we confirm deregulation of genes known to play driving roles in progression to metastasis, including ATM, APC, and PTEN, and nominated genes recurrently disrupted, including CHD1, PJA2, and ETS2, an ETS family member that is recurrently deleted as a result of the formation of the TMPRSS2/ERG fusion. In summary, through our integrative analysis, we provide the first global description of the mutational landscape of CRPC and prioritize candidates for future study.

September 29, 2011

Name: Zhaohui Qin (30 minutes)

Research Area: statistical genomics

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

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.
Name: David Golan (15 minutes)

Research Area: Random Effect Models for GWAS

Title: Estimation of Heritability in Genome Wide Studies using Random Effects Models.

Abstract

Random effects models have recently been introduced as an approach for analyzing genome wide association studies (GWAS), which allows estimation of overall heritability of traits without explicitly identifying the genetic loci responsible (Yang et al. 2010). We discuss some of the heuristics used by Yang et al. and develop a method for maximum-likelihood estimation under their framework, which does not require these heuristics. Our method provides more stable and accurate heritability estimates as well as estimates for the proportion of causal markers in the genome.
Name: Chris Glazner (15 minutes)

Research Area: Gene mapping and Linkage

Title:Improving Pedigree-based Linkage Analysis by Estimating Coancestry Among Families

Abstract

We present a method for improving the power of linkage analysis by detecting chromosome segments shared identical by descent (IBD) by individuals not known to be related. Existing Markov chain Monte Carlo methods sample descent patterns on pedigrees conditional on observed marker data. These patterns can be stored as IBD graphs, which express shared ancestry only, rather than specific family relationships. A model for IBD between unrelated individuals allows the estimation of coancestry between individuals in different pedigrees. IBD graphs on separate pedigrees can then be combined using these estimates. We report results from analyses of three sets of simulated marker data on two different pedigrees. We show that when families share a gene for a trait due to shared ancestry on the order of tens of generations, our method can detect a linkage signal when independent analyses of the families do not.

October 11, 2011

Name: Shihua ? (30 minutes)

Research Area: ?

Title: Simultaneous integration of multiple types of data to discover microRNA-gene co-modules

Abstract
Name: Jessica Li (15 minutes)

Research Area: high-throughtput sequencing data analysis

Title: Sparse Linear Modeling of RNA-seq Data for Isoform Discovery and Abundance Estimation

Abstract

Since the inception of RNA-Seq technology, various attempts have been made to utilize RNA-Seq data in assembling full-length mRNA isoforms de novo and estimating abundance of isoforms. However, for genes with more than a few exons, the problem tends to be challenging, and often involves identifiability issues in statistical modeling. We have developed a statistical method called Sparse Linear modeling of RNA-Seq data for Isoform Discovery and abundance Estimation (SLIDE) that takes exon boundaries and RNA-Seq data as input to discern the set of mRNA isoforms that are most likely to present in an RNA-seq sample. SLIDE is based on a linear model with a design matrix that models the sampling probability of RNA-Seq reads from different mRNA isoforms. To tackle the model unidentifiability issue, SLIDE uses a modified Lasso procedure for parameter estimation. Compared with deterministic isoform assembly algorithms (e.g. Cuffinks), SLIDE considers the stochastic aspects of RNA-Seq reads in exons from different isoforms and thus has increased power in detecting more novel isoforms. Another advantage of SLIDE is its flexibility of incorporating other transcriptomic data such as RACE, CAGE, and EST into its model to further increase isoform discovery accuracy. SLIDE can also work downstream of other RNA-Seq assembly algorithms to integrate newly discovered genes and exons. Besides isoform discovery, SLIDE sequentially uses the same linear model to estimate the abundance of discovered isoforms. Simulation and real data studies show that SLIDE performs as well as or better than major competitors in both isoform discovery and abundance estimation.
Name: Valerie Hower (15 minutes)

Research Area: high throughput sequencing data analysis

Title: Coverage statistics for sequence census methods

Abstract

I will introduce a way of depicting fragments from a high throughput sequencing experiment as points in the plane. Regardless of fragment length distribution, these points form a two-dimensional spatial Poisson process. I will briefly mention two applications that build on this theory--a ChIP-Seq peak caller and a statistical test to quantify the uniformity of mapped reads.

October 13, 2011

Name: Julia Roman (15 minutes)

Research Area: Evolutionary genomics - population genetics

Title: Gaussian processes based Estimation of effective population

Abstract

Changes in population size influence genetic diversity of the population and, as a result, leave imprints in genomes of individuals in the population. We are interested in an inverse problem of reconstructing past population dynamics from genomic data. We start with a standard framework based on the coalescent, a stochastic process that generates genealogies connecting randomly sampled individuals from the population of interest. These genealogies serve as a glue between the population demographic history and genomic sequences. It turns out that times at which genealogical lineages coalesce contain all information about population size dynamics. Viewing these coalescent times as a point process, estimation of population size trajectories is equivalent to estimating a conditional intensity of this point process. Therefore, our inverse problem is similar to the estimation of an inhomogeneous Poisson process intensity function. We demonstrate how recent advances in Gaussian-based nonparametric inference for Poisson process can be extended to nonparametric estimation of population size dynamics under the coalescent. We validate our method using simulated and real data and compare our Gaussian process approach to a competing Gaussian Markov random field smoothing method.
Name: Farhad Hormozdiari (15 minutes)

Research Area:

Title: Detecting the parent of origin and cis-acting gene via RNA-Seq in DBAxc57BL/6 crosses.

Abstract

In this study our goal is to detect the imprinting genes , which are effected by the parent of origin. Moreover, we will detect the cis-acting genes, which the expression level of the gene is influenced by the cross of the species. In this work RNA sample is obtained from the F1 cross between the DBA and c57BL/6. We illustrate how RNA-Seq technology can help us to detect the cis-acting and imprinting genes.
Name: Diana Low (15 minutes)

Research Area: Chromatin epigenetics and Cancer Genetics

Title: Translating the Histone Code - arginine methylations, and the use of NGS data

Abstract

Any cell in an organism derives from the same starting genetic material. Functional distinction between cell types is acquired through sequential epigenetic modifications that shape chromatin and in turn affect transcription and cellular state. DNA methylation and Post-translational modifications (PTMs) of histones convey epigenetic information that extends the coding potential of nucleic acids, a phenomenon that has been commonly referred to as “histone code”. PTMs are commonly used to modify protein function. Modifications such as phosphorylation, acetylation and methylation can influence the conformation of the modified protein and its interaction with other proteins or DNA. In the case of histones, PTMs on specific residues can influence chromatin structure and function by modifying the biochemical properties of key amino acids. Histone methylation events, especially on arginine- and lysine-residues, are among the best characterized PTMs, and many of these modifications have been linked to downstream effects. Thus far, a number of protein domains have been demonstrated to directly bind to methylated lysine residues. However, the biochemical mechanisms linking histone arginine methylation to downstream biological outputs remain poorly characterized. The work in the lab focuses on the role of histone arginine methylation in transcriptional regulation and on the crosstalk between arginine methylation and other PTMs.
Name: Emrah Kostem (15 minutes)

Research Area: Computational Genetics

Title: Increasing power of genome-wide association studies by collecting additional single-nucleotide polymorphisms

Abstract

Genome-wide association studies (GWAS) have been effectively identifying the genomic regions associated with a disease trait. In a typical GWAS, an informative subset of the single nucleotide polymorphisms (SNPs), called tag SNPs, are genotyped in case-control individuals. Once the tag SNP statistics are computed, the genomic regions that are in linkage disequilibrium (LD) with the most significantly associated tag SNPs are believed to contain the causal polymorphisms. However, such LD regions are often large and contain many additional polymorphisms. Following up all the SNPs included in these regions is costly and infeasible for biological validation. In this paper we address how to characterize these regions cost-effectively with the goal of providing investigators a clear direction for biological validation. We introduce a follow-up study approach for identifying all untyped associated SNPs by selecting additional SNPs, called follow-up SNPs, from the associated regions and genotyping them in the original case-control individuals. We introduce a novel SNP selection method with the goal of maximizing the number of associated SNPs among the chosen follow-up SNPs. We show how the observed statistics of the original tag SNPs and human genetic variation reference data such as the HapMap Project can be utilized to identify the follow-up SNPs. We use simulated and real association studies based on the HapMap data and the Wellcome Trust Case-Control Consortium to demonstrate that our method shows superior performance than the correlation and distance based traditional follow-up SNP selection approaches.

October 18, 2011

Name: Sagi Snir (30 minutes)

Research Area: Phylogenetics

Title: Using Computational Tools for Piecing Together Small Trees into the Large Tree of Life

Abstract

The reconstruction of evolutionary trees is a fundamental task in Biology. The increasing amount of available genomic sequences over thousands of taxa, gave rise to the task of large scale phylogenetic reconstruction. Since accurate reconstruction is limited to few dozens of taxa, the supertree approach, aims at accurately reconstructing small trees over overlapping taxa sets and subsequently amalgamate these trees into a tree over the full taxa set. Perhaps the simplest version of this task that is still widely applicable, yet quite challenging, is quartet based reconstruction. This problem lies at the root of many tree reconstruction methods and theoretical as well as experimental results have been reported. Nevertheless, fundamental problems such as dealing with conflicting quartet trees or even with arbitrary congruent quartet trees remains problematic. In a series of works we have developed a graph theoretically based approach for the supertree task. Our approach is based on a divide and conquer algorithm where our divide step uses a semi- definite programming (SDP) formulation of MaxCut in a graph representing relationships between the organisms. We devised an extremely fast SDP-like heuristic that allows us to extend the input data from several thousands of quartet trees over few dozens of species to tens of millions of quartet trees over thousands of species. These results are promising in the realm of large scale phylogenetic reconstruction. In the talk I will focus on the theoretical aspects of this line of research. Based on works with Satish Rao and Raphy Yuster. The talk is self-contained and requires no prior knowledge in Biology.
Name: Marc Harper (15 minutes)

Research Area: Bioinformatics

Title: Phenotype Sequencing

Abstract

Phenotype sequencing is a mathematical and statistical approach to the identification of mutations and genes causing and contributing to phenotypes resulting from the evolution of organisms subject to natural or artificial mutagenesis, as well as to the detection of rare-variants associated to particular phenotypes, through next-generation high-throughput DNA sequencing. This technique can significantly reduce the cost of identifying such genes in some cases. In this talk I present explicit results from experiments on biofuel-tolerant bacterial strains independently created through random chemical mutagenesis (effectively screening through background mutations and identifying the mutations and genes that confer tolerance), results from an analysis identifying a gene associated to a rare disease in six Korean subjects, and other experimental results if time permits. This is joint work with Chris Lee.
Name: Serghei Mangul (15 minutes)

Research Area:Computational Biotechnology

Title: Maximum Likelihood Estimation of Incomplete Genomic Spectrum from HTS Data

Abstract

High-throughput sequencing makes possible to process samples containing multiple genomic sequences and then estimate their frequencies or even assemble them. The maximum likelihood estimation of frequencies of the sequences based on observed reads can be efficiently performed using expectationmaximization (EM) method assuming that we know sequences present in the sample. Frequently, such knowledge is incomplete, e.g., in RNA-seq not all isoforms are known and when sequencing viral quasispecies their sequences are unknown. We propose to enhance EM with a virtual string and incorporate it into frequency estimation tools for RNA-Seq and quasispecies sequencing. Our simulations show that EM enhanced with the virtual string estimates string frequencies more accurately than the original methods and that it can find the reads from missing quasispecies thus enabling their reconstruction.

October 20, 2011

Name: Yi Xing (30 minutes)

Research Area: ?

Title: Genome-wide determination of a broad ESRP-regulated epithelial splicing network by RNA-Seq and SELEX-Seq

Abstract

Tissue-specific alternative splicing is achieved through the coordinated assembly of RNA binding proteins at specific sites to enhance or silence splicing at nearby splice sites. We used high throughput sequencing (RNA-Seq) to investigate the complete spectrum of alternative splicing events that are regulated by the epithelial-specific splicing regulatory proteins ESRP1 and ESRP2. In addition to numerous novel ESRP-regulated cassette exons, this analysis revealed ESRP regulation of many more complex types of alternative splicing, including alternative 3' end formation and alternative polyadenylation. To define binding motifs that mediate position dependent direct regulation of splicing by ESRP, SELEX-Seq analysis was performed, coupling traditional SELEX with high-throughput sequencing. Identification and scoring of high affinity ESRP1 binding motifs within ESRP target genes allowed generation of a high resolution RNA map that defines the genome-wide activity of the ESRPs. We also investigated the complex interplay between the ESRPs and Fox families of splicing regulators, revealing a significant overlap between these splicing networks via combinatorial regulation of large numbers of exons. These extensive analyses provide a comprehensive picture of ESRP-regulated splicing events and the position dependent functions of the ESRPs in a global post-transcriptional program required for epithelial cell formation and function.
Name: Francesca Chiaromonte (30 minutes)

Research Area: ?

Title: A genome-wide view of mutation rate co-variation using multivariate analyses
Paper

Abstract

Background: While the abundance of available sequenced genomes has led to many studies of regional heterogeneity in mutation rates, the co-variation among rates of different mutation types remains largely unexplored, hindering a deeper understanding of mutagenesis and genome dynamics. Here, utilizing primate and rodent genomic alignments, we apply two multivariate analysis techniques (principal components and canonical correlations) to investigate the structure of rate co-variation for four mutation types and simultaneously explore the associations with multiple genomic features at different genomic scales and phylogenetic distances. Results: We observe a consistent, largely linear co-variation among rates of nucleotide substitutions, small insertions and small deletions, with some non-linear associations detected among these rates on chromosome X and near autosomal telomeres. This co-variation appears to be shaped by a common set of genomic features, some previously investigated and some novel to this study (nuclear lamina binding sites, methylated non-CpG sites and nucleosome-free regions). Strong non-linear relationships are also detected among genomic features near the centromeres of large chromosomes. Microsatellite mutability co-varies with other mutation rates at finer scales, but not at 1 Mb, and shows varying degrees of association with genomic features at different scales. Conclusions: Our results allow us to speculate about the role of different molecular mechanisms, such as replication, recombination, repair and local chromatin environment, in mutagenesis. The software tools developed for our analyses are available through Galaxy, an open-source genomics portal, to facilitate the use of multivariate techniques in future large-scale genomics studies.

November 3, 2011

Name: Johanna Hardin (30 minutes)

Research Area: robust distances, clustering, outlier detection

Title: A method for generating realistic correlation matrices

Abstract

Simulating sample correlation matrices is important in many areas of statistics. Approaches such as generating normal data and finding their sample correlation matrix or generating random uniform $[-1,1]$ deviates as pairwise correlations both have drawbacks. We develop an algorithm for adding noise, in a highly controlled manner, to general correlation matrices. In many instances, our method yields results which are superior to those obtained by simply simulating normal data. Moreover, we demonstrate how our general algorithm can be tailored to a number of different correlation models. Finally, using our results with an existing clustering algorithm, we show that simulating correlation matrices can help assess statistical methodology.
Name: Anand Murugan (15 minutes)

Research Area: ?

Title: Mechanistic models of regulatory sequences from high throughput sequence data.

Abstract
Name: Emiliano Periera (15 minutes)

Research Area: genomics and metagenomics

Title: 16S rRNA sequence analysis: microorganisms diversity and evolution

Abstract

The analysis based on 16S rRNA sequences, recovered directly form the environmental samples using PCR technology, have been widely used to characterize and study the huge diversity of unculturable microorganisms. When it first appeared, this approach yielded a significant advance ecological microbiology, and in particular it revealed the existence of one of the most abundant and successful bacteria in the world, the group SAR11. This group has been found in the majority of water samples regardless of their geographic location and it has shown a great ecotype diversity. That's why 16S rRNA sequences from this group, together with the physicochemical data from the sampling site, constitutes an excellent opportunity to analyze the evolutionary and ecological precesses occurring in the microbial world. And continuous environmental gradients are a key scenario to study such processes. In this work we analyzed the diversity and evolutionary patterns within the group SAR11, based on 16S rRNA genes sampled from marine environmental gradients. Specifically we have studied the correlations between biogeographic data, diversity, evolutionary distance, and environmental parameters like oxygen, salinity and sample depth, among others. Preliminary results show that besides salinity, the phylogenetic topology is structured by the eco-regions that inhabit the different communities, the geographic distance, and the photoperiod of the year.

November 8, 2011

Name: Monica Jackson (30 minutes)

Research Area: Spatial Statistics

Title: ?

Abstract
Name: Gregory Ryslik (15 minutes)

Research Area: Next-Gen Sequencing

Title: Aggregate partDSA: a novel deletion/substitution/addition aggregate learning algorithm for recursive partitioning with variable ranking

Abstract
Name: Hugues Richard (15 minutes)

Research Area: statistical and computational methods applied to sequence data

Title 1: Fiona: an automatic strategy for read error correction

Abstract 1

Automatic read error correction approaches for next generation sequencing data can dramatically improve the performance of downstream analysis tasks, such as genome assembly or SNP calling. These approaches rely on correcting the read errors without a previous alignment to a reference genome. We propose a new stand alone read error correction method based on suffix trees and which supports substitution and indel corrections for any next generation sequencing platform. Parameters needed to perform the correction are estimated automatically. Fiona is provided with an efficient implementation with very small memory consumption that can be run on inexpensive hardware, but also supports multi-core parallelization if available.

Title 2: Sequential Monte Carlo strategy for RNA-Seq signal segmentation.

Abstract 2

Transcriptome Sequencing (RNA-Seq) permits a high precision in the inference of transcriptional landscape. Namely, RNA-Seq produces millions of reads that, once aligned to the reference genome, provide counts that reflect the expression activity. With an existing annotation, the number of reads inside genes can be used to estimate expression level. Transcript boundaries can also be identified from abrupt variations in the local abundance of reads using sliding windows. However, realistic probabilistic models that simultaneously account for transcript boundaries and expression levels are still not available to describe RNA-Seq data. This precludes the use of a parametric inference framework to estimate expression at the basepair level. We designed a strand specific model that includes the changes in expression level along the genome and randomness due to the read sampling. We then use a Sequential Monte Carlo approach to infer hidden expression level along a genome sequence from observed read counts. The SMC approach allows us to specify a reasonable model that fully exploits the complexity of RNA-Seq data.

November 10, 2011

Name: Alexander Schoenhuth (30 minutes)

Research Area: ?

Title: CLEVER - Clique-Enumerating Variation Finder

Abstract

We will present new statistical approaches for structural variation discovery which are based on exhaustively enumerating maximal cliques in the read alignment graph. We outperform state-of-the-art approaches, both in terms of recall and precision, and in particular in the repetitive regions of the genome.
Name: Hadassa Brunschwig (15 minutes)

Research Area: ?

Title: Fine-scale map of recombination rates and hotspots in the mouse genome

Abstract

Recombination events are not uniformly distributed and often cluster in narrow regions known as recombination hotspots. Several studies using different approaches have dramatically advanced our understanding of recombination hotspot regulation. Population genetic data has been used to map and quantify hotspots in the human genome. Genetic variations in recombination rates and hotspots usages have been explored in human pedigrees, mouse intercrosses, and by sperm typing. These studies pointed at the central role of the *PRDM9* gene in hotspot modulation. In this study, we used whole-genome resequencing in mouse inbred strains to estimate recombination rates across the mouse genome and identified historical hotspots. We show by simulation that populations of inbred mouse strains can be used to estimate recombination rates and identify positions of historical hotspots. Recombination hotspots were found to be enriched for the predicted binding sequences for different alleles of the PRDM9 protein, and to avoid transcription start sites (TSS). Comparing the inferred historical recombination hotspots with the recent genome-wide mapping of double-strand breaks (DSBs) in mouse sperm revealed a significant overlap, especially towards the telomeres. Our results suggest that inbred strains can be used to characterize and study the dynamics of historical recombination hotspots. They also strengthen previous findings on mouse recombination hotspots, and specifically the impact of sequence variants in *Prdm9*. **

Name: Shihao Shen (15 minutes)

Research Area: Alternative Splicing, RNA-Seq analysis

Title: MATS: Multivariate Analysis of Transcript Splicing.

Abstract

Ultra-deep RNA sequencing has become a powerful and popular approach for genome-wide analysis of pre-mRNA alternative splicing. We describe MATS (Multivariate Analysis of Transcript Splicing), a Bayesian statistical framework for flexible hypothesis testing of differential alternative splicing patterns on RNA-Seq data. MATS uses a multivariate uniform prior to model the between-sample correlation in exon splicing patterns, and a Markov chain Monte Carlo (MCMC) method coupled with a simulation-based adaptive sampling procedure to calculate the P value and false discovery rate (FDR) of differential alternative splicing. Importantly, the MATS approach is applicable to almost any type of null hypotheses of interest, providing the flexibility to identify differential alternative splicing events that match a given user-defined pattern. MATS can assess the statistical significance that the absolute difference in the transcript inclusion levels of an exon between two conditions exceeds any given user-defined threshold. It can also be used to detect exons with the extreme "switch-like" differential alternative splicing pattern. We evaluated the performance of MATS using simulated data and real RNA-Seq data sets. In the RNA-Seq analysis of alternative splicing events regulated by the epithelial-specific splicing factor ESRP1, we used MATS to analyze 256 million RNA-Seq reads on a human breast cancer cell line (MDA-MB-231) with ectopic expression of ESRP1 or an empty vector (EV) control. We selected 164 exons that covered a broad range of MATS FDR values for RT-PCR validation. We obtained a high RT-PCR validation rate of 86% for exons with a MATS FDR of <10%, demonstrating that MATS can reliably detect differential alternative splicing events. Additionally, over the full list of RT-PCR tested exons, the MATS FDR estimates matched well with the experimental validation rate. Our results demonstrate that MATS is an effective and flexible approach for detecting differential alternative splicing from RNA-Seq data.

November 22, 2011

Name: Carolin Kosiol (30 minutes)

Research Area: Evolutionary Genomics, Phylogenetics

Title: Estimating empirical codon mutation models from re-sequencing data

Abstract

Models of molecular sequence evolution are essential tools in bioinformatics: estimation of phylogenies, alignments, prediction of function, detection of selection. Empirical models summarize the pressures and mutational patterns of sequence evolution from large data sets. In this talk, I will present how we take advantage of the new re-sequencing data to estimate empirical codon models.
Name: Piotr W. Zwiernik (15 minutes)

Research Area: Algebraic statistics, statistics

Title: General Markov models from the algebraic perspective

Abstract

I am going to give a brief introduction to the study of phylogenetic tree models from the perspective of algebraic geometry. I will present my results on the models for binary data, where we managed to construct an interesting change of coordinates with a reach combinatorial structure. These, so called, tree cumulants enable us to get a deeper understanding of the inference for general Markov models, for example: identifiability, likelihood surface. some keywords: algebraic statistics, binary models, phylogenetic tree models, cumulants, tree cumulants.
Name: Peter Chi (15 minutes)

Research Area: Phylogenetic Approaches to Mapping Cell Lineages

Title: ?

Abstract

Phylogenetics can be defined as the study of evolutionary distances between species, with a typical goal of inferring a “species tree” which represents their evolutionary lineage. Alternatively, Phylogenetics can also be used to infer relationships between individual cells within a single organism, as each cell in the organism comes from a founder (i.e. the zygote), and mutations occur throughout the cell divisions that are necessary to develop and maintain the organism. A seminal work in this area was done in 1983 when Sulston et al. mapped the lineage of all 671 cells of a C. elegans through a physical inspection over its 12 hour embryogenesis. Such a task is virtually impossible in higher organisms, due to a greater number of cells, longer embryogenesis, and the fact that they are usually opaque. However, the framework of Phylogenetics as originally applied to species can be used here to infer these relationships among cell populations. This area of research has direct applications to any field which has interests in dividing cell populations, such as cancer, adult neurogenesis, tissue repair and maintenance, and stem cell research. In this talk, I will discuss current challenges, including the genotyping of single cells, and obtaining genetic data that is suitable for resolving a phylogenetic tree of cells within a single organism.
Name: Franz Baumdicker (15 minutes)

Research Area: population genetics, evolutionary genomics

Title: The infinitely many genes model for distributed bacterial genomes

Abstract

Individual bacteria from the same population often carry different genes. Existing approaches modelling this genomic diversity are based on the total number of genes (pangenome) and the gene frequency spectrum (supragenome). We extend these ideas by taking the sample genealogy into account. In population genetics the well known Infinitely Many Sites model is used to model the variation of a DNA sequence. In the same way we assume that new genes are taken from the environment and may be lost again along the ancestral lines. The presented Infinitely Many Genes Model fits well to data from cyanobacteria and provides statistical tests for neutrality.

December 6, 2011

Name: Wen-Yun Yang (15 minutes)

Research Area: Computational methods for population genetics, and high-throughput sequencing

Title: A model-based approach to analysis of spatial structure in genetic data

Abstract
Name: Aaron Kleinman (15 minutes)

Research Area: ?

Title: ?

Abstract
Name: Zhiyuan Zhai (15 minutes)

Research Area: Biostatistics

Title: Normal and Compound Poisson Approximations for Pattern Occurrences in NGS Sequence Reads

Abstract

Next generation sequencing (NGS) technologies are now widely used in many biological studies. In NGS, sequence reads are randomly sampled from the genome sequence of interest. Most computational approaches for NGS data first map the reads to the genome and then analyze the data based on the mapped reads. Since many organisms have unknown genome sequences and many reads cannot be uniquely mapped to the genomes even if the genome sequences are known, alternative analytical methods are needed for the study of NGS data. Word pattern counting, the study of the probabilistic distribution of the number of occurrences of word patterns in one or multiple long sequences, has played important roles in molecular sequence analysis. However, no studies are available on the distribution of the number of occurrences of word patterns in sequence reads from NGS. In this paper, we build probabilistic models for the background sequence and the sampling process of the sequence reads from the genome. Based on the models, we study normal and compound Poisson approximations for the number of occurrences of word patterns from the sequence reads. The study needs to consider the randomness in generating the long background sequence as well as in the sampling of the reads using NGS. We show the accuracy of these approximations under a variety of conditions for different patterns with various characteristics. Under realistic assumptions, the compound Poisson approximation seems to outperform the normal approximation in most situations. These approximate distributions can be used to evaluate the statistical significance of the occurrence of patterns from NGS data. The theory and the computational algorithm for calculating the approximate distributions are then used to analyze a Chip-Seq data using transcription factor GABP. We show that weak GABP binding motif signals are present in the control data set indicating some Chip residue effects. With the Chip-Seq data set, much stronger GABP motif signals are identified.
Name: ? (15 minutes)

Research Area: ?

Title: ?

Abstract

December 8, 2011

Name: ? (15 minutes)

Research Area: ?

Title: ?

Abstract
Name: ? (15 minutes)

Research Area: ?

Title: ?

Abstract
Name: ? (15 minutes)

Research Area: ?

Title: ?

Abstract
Name: ? (15 minutes)

Research Area: ?

Title: ?

Abstract

Mathematical and Computational Approaches in High-Throughput Genomics

Off Week Session Summary

Schedule