Large-Scale Statistical Inference of Gene Regulatory Networks: Local Network-Based Measures

Abstract

In this chapter we discuss various local network-based measures in order to assess the performance of inference algorithms for estimating regulatory networks. These statistical measures represent domain specific knowledge and are for this reason better adapted to problems that are directly involving networks compared to other measures frequently used in this context like the F-score. We are discussing three such measures with special focus on the inference of regulatory networks from expression data. However, due to the fact that currently there is a vast interest in network-based approaches in systems biology the presented measures may be also of interest for the analysis of a different type of large-scale genomics data.