Expression difference analysis of cancer and non-cancer gene based on protein interaction network

摘要

Study of human protein interaction networks, cancer and non-cancer gene differences in topology, which provides important basis for potential cancer genes develop. First to use more than one database, build a more comprehensive human protein interactions (protein-protein interaction, PPI) network, and collect a known cancer genes and extracting large numbers of random samples from PPI networks, and statistical analysis of random samples and known cancer gene in the nodes, number of nodes and the largest connected component of the degree of difference. The following results were obtained: (1) build a more comprehensive human protein interaction networks, (2) access to sets of breast cancer genes known, and (3) random sample of known cancer gene sets and PPI network nodes, node, and maximum connectivity components such as the topology properties differ significantly (p<2.15x10~(-16)). PPI network nodes, node, and maximum number of connected components such as topology attribute can contribute significantly to distinguish between cancer and non-cancer-related genes, for new and unknown cancer gene provided important investigation based on excavations.