Mining Cancer-associated Protein Sophisticated Features And Rebuild The Subnet Of Cancer-related Functions

Currently,the knowledge of cancer genes(proteins) and their associated functions or biological pathways is limited,remaining as a major hurdle to cancer studies.Specifically,many cancer proteins are only annotated to high-level GO(Gene Ontology) categories,and thus finding their finer functions is of great value for interpreting the functional mechanisms of cancer.Here,based on protein-protein interaction sub-networks defined in function-specific subspaces by using Gene Ontology,we suggest an approach to find the finer functions of cancer-associated proteins.The predicted finer functions are highly valuable for guiding the follow-up wet-lab research on the molecular mechanism of cancer.As applications of the proposed algorithms,we predict finer functions for proteins associated with OSCC(oral squamous cell carcinoma),lung cancer and colon cancer.Furthermore,although cancers are heterogeneous disorders involving diverse mutations of genes,most of them would show alterations in a few essential cellular processes.Cancer proteins might collectively contribute to cancer development through alterations of some essential functions.Therefore,discovering such cancer-associated functions and interactions among proteins involved in them is meaningful for oncogenesis mechanism exploration.By constructing function-specific protein-protein interaction sub-networks,we predict new proteins to the cancer-associated functions from higher level GO categories,leading to a more integrated protein-protein interaction sub-networks associated with cancers.These results should be of great value for uncovering functional mechanisms of cancers.