Robust estimators for expression analysis

Abstract

A hierarchy of simple models is used to design robust estimators meeting these goals for both stand alone and comparative experiments. This algorithm has been validated against an extensive panel of known spike experiments, and shows comparable performance to existing standards.