Estimating the variance when assessing the number of false discoveries in Gaussian comparisons
2012 (English)Report (Other academic)
We assume that a random vector (Z1, . . . ,ZM) is given, following a multi-variate Gaussian distribution with zero means and unit variances. FollowingStorey , we consider an estimator of the number of false discoveries (FD),i.e. the number of components |Zk| exceeding some cut-off level z, based on anestimator of M. We analyze the variance of the difference between FD and theestimated value with aid of the Mehler identity, and propose a correspondingvariance estimator, based on the Fisher transformation of a given set of empiricalcorrelations. The proposed estimator is evaluated in a simulation study. Finally,we demonstrate the applicability of the obtained estimator in relation to realobservations from a microarray experiment concerning differentially expressedgenes.
Place, publisher, year, edition, pages
Uppsala universitet, 2012. , 14 p.
U.U.D.M. report / Uppsala University, Department of Mathematics, ISSN 1101-3591 ; 2012:2
Probability Theory and Statistics
Research subject Mathematical Statistics
IdentifiersURN: urn:nbn:se:uu:diva-169404OAI: oai:DiVA.org:uu-169404DiVA: diva2:506447