Reserch On Algorithm Of Discovering Composite Functional Modules Based On Integrated Networks

Composite functional modules based on protein-protein interaction and transcriptional regulatory interaction give more insight into the complexity of cellular processes and function,which are also important for understanding the function interplay between the two interactions.However,the present studies are just for a single network or the basic unit of integrated network.So there is a huge research space in combination of two relationships to identify functional modules.In this paper,a new clustering algorithm called CRMD(Composite Modules Detection)is presented to identify composite functional modules based on protein-protein interaction and transcriptional regulatory interaction.The algorithm is applied to yeast and human datasets,composite functional modules identified by CRMD are better than other algorithms in GO enrichment and KEGG enrichment.Besides,with pathway analysis,we find an association between human composite functional module and human disease,exemplified by cancer,we find that the cancer pathway enriched modules are enriched with cancer-related genes.With those biological analysis,we prove that modules identified by CRMD are functional consistent and have more value.The algorithm considers the topological characteristics of the composite function module,with further analyzing the relationship of transcriptional regulation and protein-protein interaction,there is a special association between target genes,called C-pair,which exist in target genes.So,a new method called CPCRM to discovering composite functional modules based on C-pair is proposed.Applying CPCRM in human and yeast dataset,the result of CPCRM have a biologically significant proportion higher than CRMD's.In yeast datasets,we analyse the gene expression of nuclear genes and modules,association between protein complex and composite modules and predict new transcriptional regulation.In human dataset,based on previous studies,we confirm active composite modules associated with breast cancer.The results show that modules discovered by CPCRM enriched significant biological information.