Texas A&M - Beihang Summer Program

Date: August 2, 2019

Time: 1:15PM - 2:15PM

Location: BLOC 220

Speaker: Alan Dabney, Texas A&M University - Statistics

Description:
Title: Genomics and genome-wide statistical hypothesis testing
Abstract: Genomic data are high-dimensional, consisting of thousands-to-millions of features. One common application is to apply statistical hypothesis testing to select from the list of features a subset that appear to exhibit differences between comparison groups. For example, if the comparison groups are cancer and normal, we might want to find the subset of genes that are differentially expressed in cancer subjects relative to normal subjects. Classical statistical hypothesis testing was designed for application to one or a few hypothesis tests, and hence the classical Type I error rate and its associated p-value are not ideally applied to genomic data. Instead, an alternative error measure called the False Discovery Rate (FDR) and its associated q-value have become the standard. In this talk, I will present a brief overview of genomics data, FDR and q-values and talk through a differential expression analysis of a small real example RNA-Seq dataset of transcription measurements for 60,000 genes under two conditions.