A General Framework for Multiple Testing Dependence
Tuesday, January 27, 2009
04:30 PM in LILY G126
Professor John Storey
Department of Molecular Biology, Princeton University
I will present a general framework for performing large-scale significance testing in the presence of arbitrarily strong dependence. We have derived a low-dimensional set of random vectors, called a dependence kernel, that fully captures the dependence structure in an observed high-dimensional data set. This result shows a surprising reversal of the "curse of dimensionality" in the high-dimensional hypothesis testing setting. We have also shown theoretically that conditioning on a dependence kernel is sufficient to render statistical tests independent regardless of the level of dependence in the observed data. Finally, I demonstrate that this framework for multiple testing dependence has implications in a variety of common multiple testing problems, such as in gene expression studies, brain imaging, and spatial epidemiology. This is joint work with Jeffrey Leek.
Associated reading:
Leek JT and Storey JD (2008), A general framework for multiple testing dependence. PNAS, 105: 18718-18723.