Predicting Pathogenic Missense Mutations in the Human c-MET Oncogene Using a Nucleotide Scoring Functio

Abstract

Background: Many nucleotide variations in the human genome remain uncharacterized even more than ten years after the first draft was published. Objective: A three-parameter nucleotide-based scoring function was designed to predict the possible pathogenicity of 163 uncharacterized but validated missense mutations of the c-MET oncogene. Methodology: The parameters used by the scoring function were: surrounding consensus regions in a multiple sequence alignment of the human c-MET oncogene and five orthologous variants, inter-species allele frequency of each human missense mutation, and the nature of the mutation. Results: Out of 163 variants of unknown significance, 99 and 52 mutations were characterized as likely pathogenic and pathogenic respectively. An analysis of nine variants of known significance revealed that this scoring function could classify six out of nine (66.67%) variants with a reasonable accuracy.