Stanford CCSB Seminar Series
2010 Seminars
Reverse Engineering Tree-Evolving Gene Networks Underlying Developing Breast Cancer Cell Lineages
Tuesday November 20, 2010 at 3 pm (Li Ka Shing Center Room LK101)
Eric P. Xing, Associate Professor of Computer Sciences, Carnegie Mellon University
Estimating gene regulatory networks over biological lineages is central to a deeper understanding of how cells evolve during development and differentiation. However, one challenge in estimating such evolving networks is that their host cells are not only contiguously evolving, but also branching over time. For example, stem cells evolve into two more specialized daughter cells at each division, forming a tree of networks. Another example is in a laboratory setting: a biologist may apply several different drugs to a malignant cancer cell to analyze the changes each drug has produced in the treated cells. Each treated cell is not directly related to another treated cell, but rather to the malignant cancer cell that it was derived from. We propose a novel algorithm Treegl, which builds on the L1 plus total variation penalized graphical logistic regression to effectively estimate multiple evolving gene networks corresponding to cell types related by a tree-genealogy, based on only a few samples from each cell type. Treegl takes advantage of the similarity between related networks along the biological lineage, while at the same time exposing sharp differences between the networks. We demonstrate that our algorithm performs significantly better than existing methods via simulation. Furthermore we explore an application to a breast cancer analysis. Based on only a few microarray measurements, our algorithm is able to produce biologically valid results that provide insight into the progression and reversion of breast cancer. Joint work with Ankur P. Parikh, Wei Wu, and Ross E. Curtis
Predictive computational models for identifying genotype-specific targeted cancer therapeutics
Tuesday October 19, 2010 @ 4pm (Clark Center S363)
Adam Margolin, PhD, Biomedical Informatics, Broad Institute
Recent technological advances have enabled characterization of large tumor cell line panels with respect to dense molecular features, such as somatic mutations and gene expression profiles, and sensitivity to perturbations, measured using genome-wide RNAi or high-throughput compound screens. The emergence of such information, coupled with the rapidly sharpening window into cancer genetics provided by large-scale genome projects, now makes possible a research direction aimed at linking the constellation of mutations in a tumor cell with dysregulated cellular networks and therapeutic interventions. Such a capability would 1) improve the success rate of clinical trials and the response rate of existing therapeutics, 2) yield insights into the multi-scale cellular networks underlying tumor initiation and maintenance, 3) identify novel therapeutic targets, and 4) provide an analytical framework for predicting other cellular phenotypes. In this talk I will describe recent progress in applying regularized regression models to identify genotype-specific targeted therapeutics. I will begin by discussing a project aimed at identifying compounds that inhibit the anti-apoptotic BCL2 family protein, MCL1, together with experimentally validated genetic features that sensitize cancer cells to this class of compounds. I will then present a number of synthetic lethality predictions inferred by applying the same computational framework to two large-scale chemical screening datasets.
An Integrated Approach to Uncover drivers of Cancer
Tuesday September 14, 2010 @ 2pm (Clark Center S361)
Dana Pe'er, Assistant Professor, Biological Sciences, Columbia University
Systematic characterization of cancer genomes has revealed a staggering number of diverse aberrations that differ among individuals, such that the functional importance and physiological impact of most tumor genetic alterations remains poorly defined. We developed a computational framework that integrates chromosomal copy number and gene expression data for detecting aberrations that promote cancer progression. We demonstrate the utility of this framework using a melanoma dataset. Our analysis correctly identified known drivers of melanoma and predicted multiple novel tumor dependencies. Two dependencies, TBC1D16 and RAB27A, confirmed empirically, suggest that abnormal regulation of protein trafficking contributes to proliferation in melanoma. Together, these results demonstrate the ability of integrative Bayesian approaches to identify novel candidate drivers with biological, and possibly therapeutic, importance in cancer.
