Estimated Pathogenicity of Kinase Mutants(EPKiMu)
Description
The pathogenicity of more than 42,000 kinase mutants for 248 kinases (collected from the COSMIC) was predicted using PolyPhen-2, SIFT, PredictSNP and FATHMM and a consensus was also derived based on the above four softwares. Each file contains predicted deleterious/cancerous nature for all the mutants along with the consensus derived for all the softwares as indicated by a color (Black - Pathogenic prediction by all 4 softwares; dark grey - deleteriou s by 3 softwares or deleterious by any two softwares and cancerous by FATHMM; light grey- deleterious by any 2 softwares or deleterious by 1 software and cancerous by FATHMM). This data serves as a valuable resource for clinicians to check if a kinase mutant identified in patient's sample is a driver or a passenger and helps in designing an appropriate treatment strategy. In addition, this database may serve as a training/test dataset for bioinformaticians to develop and validate novel computational tools.
Files
Steps to reproduce
The mutations spanning 248 kinases were manually collected from the COSMIC (Catalogue of Somatic Mutations in Cancer) database. Predictions for deleterious nature of mutations with PredictSNP, PolyPhen-2 and SIFT were collected from PredictSNP server and cancerous nature of mutations were collected from FATHMM server. Data was curated and consensus for all the softwares was derived manually. Certain regions/sub-regions within each kinase were depicted, where available.