This week's Statistics Seminar Speaker will be Dr. Chiara Sabatti, an associate professor of both Health Research and Policy (Biostatistics) and Statistics at Stanford University. Dr. Sabatti completed her undergraduate and graduate studies in Statistics and Economics from Bocconi University, earned a PhD in Statistics from Stanford, and remained at Stanford to complete Post Doctoral work in Genetics. She is also a member of Stanford Bio-X, an interdisciplinary biosciences institute that specializes in the human body. For more information on Dr. Sabatti's work, including published research, please visit her webpage.
Title: Controlling the False Discovery Rate in Genome Wide Association Studies: Two Stories
Abstract: Genome wide association studies (GWAS) have become a routine tool to analyze the genetic basis of complex traits. Thousands of genomic loci have been implicated for a variety of phenotypes using this design and a fairly simple statistical analysis.
There is evidence, however, that a number of relevant associations are yet to be discovered. Within this framework, we have worked to outline analysis strategies that might improve power, while controlling the FDR of the discovery of genetic loci with phenotypic effects. I will summarize our progress in two directions: in one case we capitalize on the multiplicity of available phenotypes, in another we adapt to the multivariate nature of the genetic signature. While many contributed to this work, let me single out C. Peterson, Y. Benjamini and M. Bogdan.
Refreshments will be served after the seminar in 1181 Comstock Hall.