Analysis And Application Of The Gene Semantic Similarity Based On Disease Phenotype

In this genomics era, there has been huge amount of biological data about genes,and OMIM(Online Mendelian Inheritance In Man) is one of these. With thedevelopment of bioinformatics, computation of gene similarity and gene functionalrelation data mining become the new focus. As a kind of similarity, semantic similarityis good at representing gene functional similarity, which is based on ontology.People are paying more attentions to diseases, and searching the new methods tocure diseases has been hot topic, in the meantime, more and more researchers focusedtheir mind on disease-causing genes, hope they can find new ways to cure diseases. So,we need compute disease-causing gene semantic similarity based on disease phenotypesto explore its pathogenicity relationship, and then improve the development in diseasesfield, but now the research work about this has not been system developed, and there areno tools with this function.HPO(Human Phenotype Ontology) is an important ontology about relationshipsbetween human diseases and disease-causing genes, and its data comes from OMIM.This article did research on computation of disease-causing gene semantic similaritybased on disease phenotype, and implemented two software packages, HPO.db andHPOSim. The HPO.db package integrated data in HPO, users can query diseasephenotypes’ information; and the HPOSim package included many methods of genesemantic similarity based on disease phenotype. We proved that the functions ourpackages supplied were correct through lots of experiments.