PROJECT TITLE :

An Independent Filter for Gene Set Testing Based on Spectral Enrichment

ABSTRACT:

Gene set testing has become an imperative tool for the analysis of high-dimensional genomic knowledge. An necessary motivation for testing gene sets, instead of individual genomic variables, is to improve statistical power by reducing the quantity of tested hypotheses. Given the dramatic growth in common gene set collections, but, testing is usually performed with nearly as many gene sets as underlying genomic variables. To address the challenge to statistical power posed by giant gene set collections, we have developed spectral gene set filtering (SGSF), a unique technique for independent filtering of gene set collections previous to gene set testing. The SGSF technique uses as a filter statistic the p-price measuring the statistical significance of the association between each gene set and therefore the sample principal elements (PCs), taking into consideration the significance of the associated eigenvalues. Because this filter statistic is independent of standard gene set test statistics under the null hypothesis however dependent beneath the choice, the proportion of enriched gene sets is increased while not impacting the sort I error rate. As shown using simulated and real gene expression information, the SGSF algorithm accurately filters gene sets unrelated to the experimental outcome resulting in considerably increased gene set testing power.