Identifying differentially expressed pathways via a mixed integer linear programming model

The identification of genes and pathways involved in biological processes is a central problem in systems biology. Recent microarray technologies and other high-throughput experiments provide information which sheds light on this problem. In this article, the authors propose a new computational method to detect active pathways, or identify differentially expressed pathways via integration of gene expression and interactomic data in a sophisticated and efficient manner. Specifically, by using signal-to-noise ratio to measure the differentially expressed level of networks, this problem is formulated as a mixed integer linear programming problem (MILP). The results on yeast and human data demonstrate that the proposed method is more accurate and robust than existing approaches.