Characterizing the spread of the effect of mutations on the protein-protein interaction network

Networks can serve as a powerful tool for dissecting phenomena in complex biological systems. Of particular interest in biology is the effect of mutations on the resulting phenotype. Although numerous prior studies address this for specific cases, a generalized mechanistic theory of the phenotypic effect of mutation-induced perturbations does not exist. Here we identify patterns and principles characterizing the spread of the effect of mutations on the protein-protein interaction (PPI) network. Specifically, we propose network-based laws to identify differentially expressed genes (DEGs) following loss-of-function (LOF) mutation by studying the interplay between the topology of the PPI network and its response following node removal. We utilize cell line gene expression data following LOF mutation to evaluate network measures involving the DEGs and the knocked-out gene (KOG). We find that for 26/53 genes the DEGs form a statistically significant connected subgraph in a PPI network comprised of interactions curated from literature. Furthermore, we find that DEGs are significantly proximal to the KOG in 10/53 cases. This work contributes fundamental systems-level observations which may be exploited for the development of a predictive framework for general use in biological research.