TY - JOUR T1 - A Network-based Approach for Predicting Missing Pathway Interactions A1 - Navlakha, Saket A1 - Gitter, Anthony A1 - Bar-Joseph, Ziv Y1 - 2012/08/16 N2 - Author Summary Networks of protein interactions encode a variety of molecular processes occurring in the cell. Embedded within these networks are important subnetworks called signaling pathways. Pathways are initiated by upstream proteins (called sources) that receive signals from the environment and trigger a cascade of information to downstream proteins (targets). Modeling the interactions that occur within this cascade is important because pathway disruption has been linked to several diseases. Further, the interactions help us better understand how cells respond to various conditions and environments. Unfortunately, interaction networks today are largely incomplete, which makes this analysis difficult. We provide a framework to model missing interactions in pathways by searching for interactions that putatively result in quicker and more efficient source-target cascades. We find that we can substantially shorten source-target distances with only a few additional edges and that many of our predicted edges have support in several knowledge databases and literature reports. We believe our approach will be useful to identify interesting and important pathway-centric interactions that have been missed by previous experimental assays. JF - PLOS Computational Biology JA - PLOS Computational Biology VL - 8 IS - 8 UR - https://doi.org/10.1371/journal.pcbi.1002640 SP - e1002640 EP - PB - Public Library of Science M3 - doi:10.1371/journal.pcbi.1002640 ER -