Genetic diversity of 'Very Important Pharmacogenes' in two South-Asian populations

Abstract

Objectives: Reliable identification of population-specific variants is important for building the single nucleotide polymorphism (SNP) profile. In this study, genomic variation using allele frequency differences of pharmacologically important genes for Gujarati Indians in Houston (GIH) and Indian Telugu in the U.K. (ITU) from the 1000 Genomes Project vis-à-vis global population data was studied to understand its role in drug response.

Methods: Joint genotyping approach was used to derive variants of GIH and ITU independently. SNPs of both these populations with significant allele frequency variation (minor allele frequency ≥ 0.05) with super-populations from the 1000 Genomes Project and gnomAD based on Chi-square distribution with p-value of ≤ 0.05 and Bonferroni's multiple adjustment tests were identified. Population stratification and fixation index analysis was carried out to understand genetic differentiation. Functional annotation of variants was carried out using SnpEff, VEP and CADD score.

Results: Population stratification of VIP genes revealed four clusters viz., single cluster of GIH and ITU, one cluster each of East Asian, European, African populations and Admixed American was found to be admixed. A total of 13 SNPs belonging to ten pharmacogenes were identified to have significant allele frequency variation in both GIH and ITU populations as compared to one or more super-populations. These SNPs belong to VKORC1 (rs17708472, rs2359612, rs8050894) involved in Vitamin K cycle, cytochrome P450 isoforms CYP2C9 (rs1057910), CYP2B6 (rs3211371), CYP2A2 (rs4646425) and CYP2A4 (rs4646440); ATP-binding cassette (ABC) transporter ABCB1 (rs12720067), DPYD1 (rs12119882, rs56160474) involved in pyrimidine metabolism, methyltransferase COMT (rs9332377) and transcriptional factor NR1I2 (rs6785049). SNPs rs1544410 (VDR), rs2725264 (ABCG2), rs5215 and rs5219 (KCNJ11) share high fixation index (≥ 0.5) with either EAS/AFR populations. Missense variants rs1057910 (CYP2C9), rs1801028 (DRD2) and rs1138272 (GSTP1), rs116855232 (NUDT15); intronic variants rs1131341 (NQO1) and rs115349832 (DPYD) are identified to be 'deleterious'.

Conclusions: Analysis of SNPs pertaining to pharmacogenes in GIH and ITU populations using population structure, fixation index and allele frequency variation provides a premise for understanding the role of genetic diversity in drug response in Asian Indians.

Keywords: 1000 Genomes Project; Allele frequency; GIH; ITU; Pharmacogenes; SNPs; Variant calling; gnomAD.

Figure 1. Flow-chart for identification and analysis of SNPs pertaining to VIP genes.

Figure 2. (A) Venn diagram depicting the common and unique SNPs belonging to VIP genes identified by joint genotyping of GIH population and that from 1KGP. (B) Venn diagram depicting the common and unique SNPs belonging to VIP genes identified by joint genotyping.

Figure 3. Genetic diversity of VIP genes.

(A) Population stratification of VIP genes at k = 4; (B) Principal Component analysis of VIP genes. Color legend: AFR in dark blue, EUR in lavender, EAS in olive green, AMR1 in mustard, AMR2 in beige, GIH in black and ITU in stone grey.

Figure 4. Venn diagram depicting SNPs with significant allele frequency variation in GIH and ITU with other populations/super-populations in 1KGP and gnomAD.

Figure 5. Venn diagram depicting shared and unique SNPs in GIH and ITU.

Figure 6. Histogram of SNPs (shared by GIH and ITU) belonging to VIP genes that show significant allele frequency variation with at least one super-population from 1KGP.

Figure 7. Histogram of SNPs in GIH population belonging to VIP genes that show significant allele frequency variation with at least one super-population from 1KGP.

Figure 8. Histogram of SNPs in ITU population belonging to VIP genes that show significant allele frequency variation with at least one super-population from 1KGP.