David Geffen School of Medicine at UCLA
Department of Human Genetics

Speaker Series - Winter Quarter 2013

Mondays, 11am - 12pm, Gonda Building First Floor Conference Room, 1357

Mon, Jan 14
Why the genetic heterogeneity hypothesis for complex trait variation deserves a second look
Anthony D Long , Ph.D., Professor, Ecology & Evolutionary Biology, University of California, Irvine
Contact & Intro: Marc Suchard, msuchard@ucla.edu
Thu, Jan 31
Geology 3656 at 12:30pm
Statistical Methods for Analyzing High-throughput Genomic Data
Jingyi Jessica Li, Ph.D., Department of Statistics, UC Berkeley
Contact & Intro: Janet Sinsheimer, janets@mednet.ucla.edu
View details »

ABSTRACT: In the burgeoning field of genomics, high-throughput technologies (e.g. microarrays, next-generation sequencing and label-free mass spectrometry) have enabled biologists to perform global analysis on thousands of genes, mRNAs and proteins simultaneously. Extracting useful information from enormous amounts of high-throughput genomic data is an increasingly pressing challenge to statistical and computational science. In this talk, I will present three projects in which statistical and computational methods were used to analyze high-throughput genomic data to address important biological questions. The first part of my talk will demonstrate the power of simple statistical analysis in correcting biases of large-scale protein level estimates and in understanding the relationship between gene transcription and protein levels. The second part will focus on a statistical method called "SLIDE" that employs probabilistic modeling and L1 sparse estimation to answer an important question in genomics: how to identify and quantify mRNA products of gene transcription (i.e, isoforms) from next generation RNA sequencing data? In the final part, I will introduce an ongoing project where we developed a new statistical measure under a local regression and clustering framework to capture non-functional relationships between a pair of variables. This new measure will have broad potential applications in genomics and other fields.

Mon, Feb 04
Regulation of TGF-beta signal transduction by mono- and deubiquitylation of Smads
Stuart J. Newfeld, Ph.D., Professor, Faculty Leader of Cellular and Molecular Biosciences, Director of the IMSD at ASU, School of Life Sciences, Arizona State University
Contact & Intro: Julian Martinez, x42405
Thu, Feb 07
Geology 3656 at 1pm
Detection of Cancer Subgroup Associated Alternative Splicing
Jianhua Hu, Ph.D., Associate Professor, Department of Biostatistics, University of Texas M.D. Anderson Cancer Center
Contact & Intro: Janet Sinsheimer, janets@mednet.ucla.edu
View details »

ABSTRACT: Alternative splicing is known to be a critical factor in cancer formation and progression and can be detected using Affymetrix exon tiling arrays. An important observation is that alternative splicing events may show differential degrees among subgroups of cancer patients, and those subgroups could have common clinical characteristics and are therefore biologically meaningful. We propose modeling exon tiling array data by a structured Gaussian mixture model, integrated with low-rank approximation of the interaction space based on ANOVA-type representation. We develop a sequential test for detecting possible cancer subgroup structure and possible subgroup-associated alternative splicing events. For detecting the existence of subgroups, we use a penalized likelihood ratio test. The validity and applicability of the proposed method will be demonstrated through asymptotic analysis, simulation studies, and a brain cancer study.

Thu, Feb 14
Geology 3656 at 12:30pm
Bayesian Variable Selection for High-Dimensional Genetic Association Studies
Melanie Quintana, Ph.D., Division of Biostatistics, Department of Preventive Medicine, USC
Contact & Intro: Janet Sinsheimer, janets@mednet.ucla.edu
View details »

ABSTRACT: Recent advances in genotyping technology have resulted in a dramatic change in the way hypothesis-based genetic association studies are conducted. Previous investigators who were limited by cost and power to investigate only a handful of common variants within the most interesting genes are now able to conduct whole genome studies involving millions of both common and rare genetic variants. Thus, the bottleneck in understanding many important complex diseases has become how to sift through millions of exchangeable variants. While the most common approach to this problem has been to apply a marginal test to all genetic markers, the analytical strategies that are a focus of my research aim to improve upon these methods by modeling the outcome variable as a function of a multivariate genetic profile using Bayesian model uncertainty techniques. Throughout this talk I will demonstrate that these techniques can be extremely powerful and advantageous in high-dimensional variable selection problems due to their flexibility and ability to provide formal intuitive multi-level inference. In particular, I will describe several methods that have been developed to gain insight into the genetic contribution (of both common and rare variants) of various complex diseases within the North Carolina Ovarian Cancer Study (NCOCS), Women’s Environmental Cancer and Radiation Epidemiology (WECARE) Study, and the Pharmacogenetics of Nicotine Addiction and Treatment Consortium (PNAT). While this work is motivated by a very specific application, I will also highlight several areas of general methodological development throughout in efficient model search algorithms and multiplicity corrected model space priors. Finally, I will describe a recent extension of this work that allows the incorporation of external variant specific biological information to guide the variable selection procedure.

Thu, Feb 21
Geology 3656 at 12:30pm
Graph-based change-point detection
Hao Chen, Ph.D., Department of Statistics, Stanford University
Contact & Intro: Janet Sinsheimer, janets@mednet.ucla.edu
View details »

ABSTRACT: After observing snapshots of a network, can we tell if there has been a change in dynamics? After reading chapters of a historical text, can we tell if there has been a change in authorship? Given a sequence of independent observations, we are concerned with testing the null hypothesis of homogeneity versus change-point alternatives, where a segment of the sequence differs in distribution from the rest. This problem has been well studied for observations in low dimension. Currently, many problems can be formulated in the change-point framework but with observations that are high-dimensional or non-Euclidean, where existing methods are limited. We develop a general nonparametric framework for change-point detection that relies on a similarity measure on the sample space of observations. This new approach, which relies on graph-based tests, can be applied to high dimensional data, as well as data from non-Euclidean sample spaces. An analytic approximation for the false positive error probability is derived and shown to be reasonably accurate by simulation. We illustrate the method through the analysis of a phone-call network from the MIT Reality Mining project and of the authorship debate of a classic western novel.

Mon, Feb 25
NRB Auditorium
Statistical and Computational Methods for Disease Mapping in Admixed Populations
Bogdan Pasaniuc, Ph.D., Assistant Professor, Institute for Molecular Medicine, Pathology & Laboratory Medicine, Geffen School of Medicine at UCLA
Contact & Intro: Marc Suchard, msuchard@ucla.edu
View details »

ABSTRACT: The genome of admixed individuals, such as African Americans or Latinos, is a mosaic of chromosomal regions originating from ancestral populations. Knowledge about the ancestral origin of each of these regions is a key component in disease mapping in admixed populations, both in correcting for false positives as well as in increasing power for novel locus discovery. In this talk, I will present novel methods for disease mapping in African Americans and discuss their limitations to more complex admixed populations such as Latinos.

Mon, Mar 11
NRB Auditorium
Fine-mapping complex trait loci
Soumya Raychaudhuri, MD, Ph.D., Assistant Professor of Medicine, Harvard Medical School Divisions of Genetics & Rheumatology Department of Medicine, Brigham and Women's Hospital
Contact & Intro: Eleazar Eskin, x51322
View details »

ABSTRACT: While genome-wide association studies have been incredibly successful at identifying loci for scores of complex traits and diseases, the causal variants that are driving disease and their mechanisms remain unknown. Here we will describe strategies taken by our group to define the causal coding and regulatory variants. We give specific recent examples from the major histocompatability locus in rheumatoid arthritis and the CFH locus in age-related macular degeneration. Finally, we illustrate how epigenetic histone modifications can be assayed and used to identify cell-specific regulatory variation that drives disease associations.

Previous Quarters