Dissecting Cancer Genes And Their Functional Cooperation Based On Gene Functional Modules

In biological system, proteins encoded by genes perform functions by interacting with each other in a modular fashion. Studying cancer genes and cancer mechanism based on gene functional modules provides us with the opportunity to dissect the cancer genes and their functional cooperation and explore cancer mechanism from a more global and systematical perspective.The main contributions are as follows:1. Studying the evolutionary conservation of cancer genes based on co-evolving gene functional modules. Functionally related proteins tend to perform functions by interacting with each other in a modular fashion, which may affect both the mode and tempo of their evolution. In this dissertation, we firstly searched for co-evolving protein protein interacting (PPI) subnetworks in the human PPI network. Identified at a given co-evolving level, we selected the subnetworks with non-randomly large sizes and defined them as co-evolving gene functional modules. Our results showed that proteins within modules tend to be conserved, evolutionarily old and enriched with proteins encoded by housekeeping genes expressed in all tissues, while proteins outside modules tend to be less-conserved, evolutionarily younger and enriched with proteins encoded by genes expressed in specific tissues. Further studying the evolutionary conservation of cancer genes based on co-evolving gene functional modules, we found that the overall conservation of cancer genes should be mainly attributed to the cancer genes encoding proteins enriched in the conserved modules. The cancer genes encoding proteins outside modules are relatively less conserved. Thus, cancer genes encoding proteins within and outside modules might play different roles in carcinogenesis, providing a new hint for studying the mechanism of cancer.2. Find co-mutated genes and candidate cancer genes in cancer genomes. The non-randomness of the co-mutation of genes in cancer samples can provide important information on the functional cooperation of gene mutations. In the current high-throughput somatic mutation screening studies, due to the relatively small sample sizes used and the extraordinary large-scale hypothesis tests, the statistical power of finding co-mutated gene pairs is very low. Based on two datasets of somatic mutations in cancer genomes, we proposed a stratified false discovery rate control approach for identifying significantly co-mutated gene pairs and candidate cancer genes. Compared with the approach of pre-selecting genes with higher mutation frequencies, many more co-mutated gene pairs and candidate cancer genes could be found by a stratified false discovery rate control strategy.3. Identifying cancer related functional modules and their interactions based on cellular location information. A Gene Ontology (GO) biological process category may encompass the genes involved in distinct processes occurring in different cellular compartments. Furthermore, the genes even within a same process may show a clear expression distinction with respect to their cellular localizations. Therefore, in this dissertation, we identified cancer related functional modules by the approach proposed by us that integrating both biological process and cellular component information. Applying this method to two cancer expression datasets, the results showed that controlling the same false discovery rate level, more and specific cancer related processes could be found by using the cellular location information. The modular network constructed by connecting pairs of cancer related signature functional modules based on PPI and gene expression data suggests a clue for further exploring the synergic action of the signature functional modules during tumorigenesis.In conclusion, based on gene functional modules, we dissected cancer genes and their functional cooperation from the aspects of sequence conservation, somatic mutation and gene expression in this dissertation.