Network-informed gene ranking tackles genetic heterogeneity in exome-sequencing studies of monogenic disease

Whole-exome sequencing studies of monogenic diseases can be hindered by genetic heterogeneity because different genes can harbour the disease-causing mutation in affected individuals. We show that interaction networks can help overcome genetic heterogeneity because many genes responsible for the same or very similar monogenic disease phenotypes interact. We present HetRank, a flexible gene-ranking method which incorporates network information. Using simulated exome sequencing studies we demonstrate that HetRank performs well under high levels of genetic heterogeneity and improves on previous efforts to overcome heterogeneity in monogenic disease studies. We make available an implementation of HetRank via the website http://sourceforge.net/p/hetran