A Novel Approach for Pathway Inference Based on Network Flow

Xianwen Ren,Xiang-Sun Zhang

2009-01-01

Abstract:Signal transduction pathways play important roles in various biological processes such as cell cycle, apoptosis, proliferation, differentiation and responses to the external stimuli. Efficient computational methods are of great demands to map signaling pathways systematically based on the interactome and microarray data in the post-genome era. In this study, we propose a novel approach to infer the pathways based on the network flow well studied in the operation research. We define a potentiality variable for each protein to denote the extent to which it belongs to the objective pathway. And the capacity on each edge is not a constant but a function of the potentiality variables of the corresponding two proteins. The total potentiality of all proteins is given an upper bound. The approach is formulated to a linear programming model and solved by the simplex method. Experiments on the yeast sporulation data suggest this novel approach recreated successfully the backbone of the MAPK signaling pathway with a low upper bound of the total potentiality. By increasing the upper bound, the approach successfully predicted all the members of the MAPK pathway responding to the pheromone. It also included the cross-talks between the MAPK pathway and the pathway controlling cell cycle which play important roles in the yeast sporulation. This simple but effective approach provides a very useful tool facilitating the biologists to mine the biological insight from the genomic data.