Computational Biology and Bioinformatics Discussion Group (3)
Thursday, May 18, 2006
Presented by the Computational Bio & Bioinformatics Discussion Group
Chris Sander, Memorial Sloan Kettering Cancer Center, "Brief Introduction to Computational Systems Biology."
Adam Olshen, Memorial Sloan Kettering Cancer Center, "A Faster Circular Binary Segmentation Algorithm and its Application to Expression Data."
Robert Klein, Rockefeller University, "Genome-wide Association Studies: Practice and Theory."
Roy Kishony, Harvard Medical School, "Epistasis Interaction Networks."
"A Faster Circular Binary Segmentation Algorithm and its Application to Expression Data"
Cancer progression often involves alterations in DNA sequence copy number. Multiple microarray platforms now facilitate high-resolution copy number assessment of entire genomes in single experiments. This technology is generally referred to as array comparative genomic hybridization (array CGH). I will discuss our method of identifying regions of abnormal copy number in array CGH data, which is called circular binary segmentation (CBS). Early versions of CBS were criticized for being slow. I will present our recently developed computational methods to greatly speed up the procedure. CBS, as well as other algorithms, have been shown to be effective for analyzing array CGH data. The second part of the talk will address the more challenging problem of identifying abnormal copy number regions in expression data.
"Genome-wide Association Studies: Practice and Theory"
Population-based genome-wide association studies are thought to have greater power for finding the genetic variants underlying common, complex disease than traditional family-based linkage studies. However, due to the large number of markers that would need to be genotyped, only recently have such studies become technically feasible. We recently undertook a genome-wide association study to identify genetic variants associated with age-related macular degeneration. We identified a tyrosine to histidine polymorphism in complement factor H that is strongly associated with this disease. I have also developed a method for computing the power of these studies when you know what SNPs you will be genotyping. Taken together, these results suggest that genome-wide association studies are a promising new approach in human genetics.
"Epistasis Interaction Networks"
Complex biological functions are encoded by networks of interacting genetic components. Epistasis, describing the way multiple perturbations (mutations or drugs) in such networks affect each other's phenotypic consequences, provides essential information for elucidating the network functional architectures. I will describe a combined experimental-theoretical approach to quantify epistatic interactions in bacteria and yeast and for using epistasis information to identify functional gene modules and their system-level organization.