The Construction Of Perturbed Functional Classes Network Of Diseases On The Basis Of Multi-omics Data

The traditional way of studying network model for systems biology is to regard the gene or protein as network node element and then to predict the function of genes or proteins through the method of clustering and classification of functional modules. While in this study, focusing on a higher level, regarding functional class as a network node, we construct functional class disturbance network model of human diseases to study the pathogenesis of the diseases.The article first introduces the idea and method of constructing the perturbed functional classes network, including the selection and processing of omics data, enrichment, the selection of perturbed functional classes and the building of network model.Then, in order to study the disturbance of functional classes of major diseases on the basis of the expression profile level, we construct perturbed functional classes network of four diseases on the basis of the expression profile level comprising cancer, neurological diseases, infectious diseases and metabolic diseases. And then we analyze the disturbance of functional classes and the relationship of the four diseases.Finally, we construct the perturbed functional classes network of cancer integrating a variety of omics data. To establish a reliable network, we select somatic gene mutations data, Methylation data, SNP data and MicroRNA data in addition to expression profiling data. By the analysis of this model, we find several pathways, including the complement, coagulation cascades, tryptophan metabolism closely linked with cancer, with little attention paid by researchers before. At last we analyze prostate cancer carcinogenesis process through the model.