A bioinformatics approach for the phenotype prediction of nonsynonymous single nucleotide polymorphisms in human cytochromes P450.

Nonsynonymous single nucleotide polymorphisms (nsSNPs) in coding regions that can lead to amino acid changes may cause alteration of protein function and account for susceptivity to disease and altered drug response. Identification of deleterious nsSNPs from tolerant nsSNPs is important for characterizing the genetic basis of human disease, assessing individual susceptibility to disease, understanding the pathogenesis of disease, identifying molecular targets for drug treatment, and conducting individualized pharmacotherapy. Numerous nsSNPs have been found in genes coding for human cytochromes P450 (P450s), but there is poor knowledge on the relationship between the genotype and phenotype of nsSNPs in P450s. We have identified 791 validated nsSNPs in 57 validated human CYP genes from the National Center for Biotechnology Information Database of Single Nucleotide Polymorphism and Swiss-Prot database. Using the polymorphism phenotyping (PolyPhen; http://genetics.bwh.harvard.edu/pph) and sorting intolerant from tolerant (SIFT; http://blocks.fhcrc.org/sift/SIFT.html) algorithms, 39 to 43% of nsSNPs in CYP genes were predicted to have functional impacts on protein function. There was a significant concordance between the predicted results using the SIFT and PolyPhen algorithms. A prediction accuracy analysis found that approximately 70% of nsSNPs were predicted correctly as damaging. Of nsSNPs predicted as deleterious, the prediction scores by the SIFT and PolyPhen algorithms were significantly associated with the numbers of nsSNPs with known phenotype confirmed by benchmarking studies, including site-directed mutagenesis analysis and clinical association studies. These amino acid substitutions are supposed to be the pathogenetic basis for the alteration of P450 enzyme activity and the association with disease susceptivity. This prediction analysis of nsSNPs in human CYP genes would be useful for further genotype-phenotype studies on individual differences in drug clearance and clinical response.