A Modified Approach in Statistical Significance for Genome Wide Studies
Document Type
Lecture
Publication Date
10-15-2010
Abstract
Most microarray studies are focused on identifying genes that are differentially expressed between two experimental conditions. One of the well-known measures of significance for genome-wide studies is the false discovery rate. The q value is estimated from a list of p values based on a permutation test for equality of means. Permutation test does not control Type I error rate to its nominal value when variances are unequal. Therefore, permutation test can be conservative or liberal. In this study, we propose a modified t-test based on jackknife estimator of population variance. The resulting p values are used to estimate q value as given by Storey and Tibshirani (2003). Monte Carlo method is used to show that the estimated q value from the proposed test controls q-value threshold better than the q value estimated from permutation test based p values.
Relational Format
presentation
Recommended Citation
Singh, Pradeep, "A Modified Approach in Statistical Significance for Genome Wide Studies" (2010). Probability & Statistics Seminar. 44.
https://egrove.olemiss.edu/math_statistics/44
