Identification of genetic polymorphisms in 40 chemotherapy pathway genes

Abstract

Background We set out to identify functionally significant common polymorphisms in 40 genes involved in the pharmacokinetic or pharmacodynamic pathways of anticancer drugs. Methods We used a pooled approach to resequence 120 DNA samples from African-Americans, Asians, and Caucasians (40 samples each). Results Approximately 338 kb of sequence was analyzed from each sample. 193 of the 711 variant loci that were identified were located in exons. The PolyMAPr program was used to mine the dbSNP, JSNP, and CGAP databases. 243 of the 711 variants were novel, and 78% of the remaining 468 were listed as unvalidated in the databases. Based on estimated allele frequencies, 274 variants were population-specific and 284 were common to all 3 groups. Of the 115 ORF SNPs, 61 were population-specific and 31 were in all 3 groups. Resequencing and database mining identified 523 variants in the ORF. The PolyPhen program was used to predict functional significance of 275 nonsynonymous biallelic SNPs: 174, 43, and 48 were predicted to be benign, possibly damaging, and probably damaging, respectively. Estimated allele frequency and BLOSUM62 score decreased with increasing predicted severity of the amino acid change. Conclusions These results will aid future pharmacogenetic studies of anticancer drugs. Clinical Pharmacology & Therapeutics (2005) 77, P63–P63; doi: 10.1016/j.clpt.2004.12.132