A Protein-Phenotype Mutual Information Based Identification of Human Disease Genes

On the basis of the assumption that diseases with similar phenotypes are caused by functionally related genes, We describe a candidate gene prioritization method that is entirely based on mutual information of human protein-protein interactions and disease phenotypic similarities, and develop a tool named GENEDIG to calculate the mutual information concordance score between the human protein network and the phenotype network. This method is applicable to genetically uncharacterized phenotypes, and also extendable to explore gene cooperativity in complex diseases. The implied association between disease and induced genes can be effectively revealed so as to facilitate future discovery of disease genes.