Detection-theoretic analysis of MatInspector

The goal of this paper is twofold. First, we wish to raise awareness in the signal processing community of research problems related to transcription factor binding site (TFBS) finding. This class of problems is important in the study of gene expression and regulation and is similar to the problems studied in statistical signal processing and detection theory. Second, we analyze one of the very popular tools for TFBS finding-MatInspector-from a detection-theoretic viewpoint. They show that MatInspector's test statistic is suboptimal, but its performance is reasonably close to that of the optimal Neyman-Pearson detector.