Identifying Ovarian Cancer Chemotherapy Response Relevant Gene Cliques

Abstract

Operation with adjuvant chemotherapy is still the principal means to treat Ovarian cancer. Identifying Ovarian Cancer Chemotherapy Response (OCCR) relevant genes and describe their interactions is thus an important issue. However the problems of high dimensional micro array data and the scarcity of biological priors make building a complete OCCR biological network intractable. To this end, we combine liquid association (LA) algorithm with biological knowledgebase searching to identify OCCR relevant gene clique and describe their interactions. Rather than trying to build a gene network, our approach focus on identifying OCCR relevant gene cliques and then patching them up. Statistical analysis and biological validation show that the identified gene cliques play important roles in tumor genesis, immunity, cells proliferation and migration etc and significantly OCCR relevant. More importantly, the connection of independent gene cliques is established and the associations of genes are described. Methodologically, the proposed method avoids the problem of complex computation, relies only on available biological priors and provides a novel way to build gene network.