PROJECT TITLE :

Graphical Features of Functional Genes in Human Protein Interaction Network

ABSTRACT:

With the completion of the human genome project, it is possible to analyze massive-scale human protein interaction network (HPIN) with complex networks theory. Proteins are encoded by genes. Essential, viable, disease, conserved, housekeeping (HK) and tissue-enriched (TE) genes are useful genes, which are organized and functioned via interaction networks. Based mostly on up-to-date data from various databases or literature, two giant-scale HPINs and 6 subnetworks are made. We have a tendency to illustrate that the HPINs and most of the subnetworks are sparse, tiny-world, scale-free, disassortative and with hierarchical modularity. Among the six subnetworks, essential, disease and HK subnetworks are additional densely connected than the others. Statistical analysis on the topological structures of the HPIN reveals that the lethal, the conserved, the HK and also the TE genes are with hallmark graphical options. Receiver operating characteristic (ROC) curves indicate that the essential genes can be distinguished from the viable ones with accuracy as high as virtually seventy%. Closeness, semi-native and eigenvector centralities will distinguish the HK genes from the TE ones with accuracy around eighty twopercent. Furthermore, the Venn diagram, cluster dendgrams and classifications of disease genes reveal that some categories of disease genes are with hallmark graphical features, especially for cancer genes, HK disease genes and TE disease genes. The findings facilitate the identification of some functional genes via topological structures. The investigations shed some light on the characteristics of the compete interactome, that have potential implications in networked medication and biological network control.