Limits...
Analysis of pharmacogenomic variants associated with population differentiation.

Yeon B, Ahn E, Kim KI, Kim IW, Oh JM, Park T - PLoS ONE (2015)

Bottom Line: Among the 17 genes related to cell communication identified in the HD gene group, five genes (STX4, PPARD, DCK, GRIK4, and DRD3) contained single nucleotide polymorphisms with Fst values greater than 0.5.In the analysis using DR genes as the background, the HD gene group contained six significant terms.Our analysis suggests that the HD gene group from PharmGKB is associated with cell communication and drug binding.

View Article: PubMed Central - PubMed

Affiliation: Interdisciplinary Program in Bioinformatics, Seoul National University, Gwanak-ro, Gwanak-gu, Seoul, Korea.

ABSTRACT
In the present study, we systematically investigated population differentiation of drug-related (DR) genes in order to identify common genetic features underlying population-specific responses to drugs. To do so, we used the International HapMap project release 27 Data and Pharmacogenomics Knowledge Base (PharmGKB) database. First, we compared four measures for assessing population differentiation: the chi-square test, the analysis of variance (ANOVA) F-test, Fst, and Nearest Shrunken Centroid Method (NSCM). Fst showed high sensitivity with stable specificity among varying sample sizes; thus, we selected Fst for determining population differentiation. Second, we divided DR genes from PharmGKB into two groups based on the degree of population differentiation as assessed by Fst: genes with a high level of differentiation (HD gene group) and genes with a low level of differentiation (LD gene group). Last, we conducted a gene ontology (GO) analysis and pathway analysis. Using all genes in the human genome as the background, the GO analysis and pathway analysis of the HD genes identified terms related to cell communication. "Cell communication" and "cell-cell signaling" had the lowest Benjamini-Hochberg's q-values (0.0002 and 0.0006, respectively), and "drug binding" was highly enriched (16.51) despite its relatively high q-value (0.0142). Among the 17 genes related to cell communication identified in the HD gene group, five genes (STX4, PPARD, DCK, GRIK4, and DRD3) contained single nucleotide polymorphisms with Fst values greater than 0.5. Specifically, the Fst values for rs10871454, rs6922548, rs3775289, rs1954787, and rs167771 were 0.682, 0.620, 0.573, 0.531, and 0.510, respectively. In the analysis using DR genes as the background, the HD gene group contained six significant terms. Five were related to reproduction, and one was "Wnt signaling pathway," which has been implicated in cancer. Our analysis suggests that the HD gene group from PharmGKB is associated with cell communication and drug binding.

No MeSH data available.


Related in: MedlinePlus

Histogram of sample sizes from 654 drug-related SNPs.A. Total sample sizes of SNPs. B. Sample size of each population of SNPs. CHB and JPT are plotted separately according to the format of the original HapMap Data. SNPs with larger sample sizes are included in Phase III, and SNPs with smaller sample sizes are included in Phase II.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4373713&req=5

pone.0119994.g001: Histogram of sample sizes from 654 drug-related SNPs.A. Total sample sizes of SNPs. B. Sample size of each population of SNPs. CHB and JPT are plotted separately according to the format of the original HapMap Data. SNPs with larger sample sizes are included in Phase III, and SNPs with smaller sample sizes are included in Phase II.

Mentions: We analyzed SNP data from International HapMap Phase II + III, release 27 (http://www.hapmap.org) [9,31]. According to Hapmap consortium, there are distinct three clusters, European, African and Asian from the principal component plot of 11 populations in hapmap3[31]. Therefore, we used three groups based on these three regions. We included 120 Yoruba from Ibadan, Nigeria (YRI), 181 Asians of which are 91 Japanese from Tokyo, Japan (JPT) and 90 Han Chinese from Beijing, China (CHB), and 120 Utah residents with ancestry from northern and western Europe (CEU). We used only founders of CEU and YRI to exclude the related samples. Because International HapMap release 27 consists of a mixture of two phases, each SNP had a different sample size. Fig. 1 shows the sample-size distributions of subpopulations from International HapMap Data. The SNPs from Phase II had smaller sample sizes, while those from Phase III had larger sample sizes (Fig. 1B). Some SNPs are only genotyped in phase II and others are only genotyped in phase III. In this reason, we only included four populations, which are both in phase II, and III simultaneously to avoid the potential biases due to different settings of each phases.


Analysis of pharmacogenomic variants associated with population differentiation.

Yeon B, Ahn E, Kim KI, Kim IW, Oh JM, Park T - PLoS ONE (2015)

Histogram of sample sizes from 654 drug-related SNPs.A. Total sample sizes of SNPs. B. Sample size of each population of SNPs. CHB and JPT are plotted separately according to the format of the original HapMap Data. SNPs with larger sample sizes are included in Phase III, and SNPs with smaller sample sizes are included in Phase II.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4373713&req=5

pone.0119994.g001: Histogram of sample sizes from 654 drug-related SNPs.A. Total sample sizes of SNPs. B. Sample size of each population of SNPs. CHB and JPT are plotted separately according to the format of the original HapMap Data. SNPs with larger sample sizes are included in Phase III, and SNPs with smaller sample sizes are included in Phase II.
Mentions: We analyzed SNP data from International HapMap Phase II + III, release 27 (http://www.hapmap.org) [9,31]. According to Hapmap consortium, there are distinct three clusters, European, African and Asian from the principal component plot of 11 populations in hapmap3[31]. Therefore, we used three groups based on these three regions. We included 120 Yoruba from Ibadan, Nigeria (YRI), 181 Asians of which are 91 Japanese from Tokyo, Japan (JPT) and 90 Han Chinese from Beijing, China (CHB), and 120 Utah residents with ancestry from northern and western Europe (CEU). We used only founders of CEU and YRI to exclude the related samples. Because International HapMap release 27 consists of a mixture of two phases, each SNP had a different sample size. Fig. 1 shows the sample-size distributions of subpopulations from International HapMap Data. The SNPs from Phase II had smaller sample sizes, while those from Phase III had larger sample sizes (Fig. 1B). Some SNPs are only genotyped in phase II and others are only genotyped in phase III. In this reason, we only included four populations, which are both in phase II, and III simultaneously to avoid the potential biases due to different settings of each phases.

Bottom Line: Among the 17 genes related to cell communication identified in the HD gene group, five genes (STX4, PPARD, DCK, GRIK4, and DRD3) contained single nucleotide polymorphisms with Fst values greater than 0.5.In the analysis using DR genes as the background, the HD gene group contained six significant terms.Our analysis suggests that the HD gene group from PharmGKB is associated with cell communication and drug binding.

View Article: PubMed Central - PubMed

Affiliation: Interdisciplinary Program in Bioinformatics, Seoul National University, Gwanak-ro, Gwanak-gu, Seoul, Korea.

ABSTRACT
In the present study, we systematically investigated population differentiation of drug-related (DR) genes in order to identify common genetic features underlying population-specific responses to drugs. To do so, we used the International HapMap project release 27 Data and Pharmacogenomics Knowledge Base (PharmGKB) database. First, we compared four measures for assessing population differentiation: the chi-square test, the analysis of variance (ANOVA) F-test, Fst, and Nearest Shrunken Centroid Method (NSCM). Fst showed high sensitivity with stable specificity among varying sample sizes; thus, we selected Fst for determining population differentiation. Second, we divided DR genes from PharmGKB into two groups based on the degree of population differentiation as assessed by Fst: genes with a high level of differentiation (HD gene group) and genes with a low level of differentiation (LD gene group). Last, we conducted a gene ontology (GO) analysis and pathway analysis. Using all genes in the human genome as the background, the GO analysis and pathway analysis of the HD genes identified terms related to cell communication. "Cell communication" and "cell-cell signaling" had the lowest Benjamini-Hochberg's q-values (0.0002 and 0.0006, respectively), and "drug binding" was highly enriched (16.51) despite its relatively high q-value (0.0142). Among the 17 genes related to cell communication identified in the HD gene group, five genes (STX4, PPARD, DCK, GRIK4, and DRD3) contained single nucleotide polymorphisms with Fst values greater than 0.5. Specifically, the Fst values for rs10871454, rs6922548, rs3775289, rs1954787, and rs167771 were 0.682, 0.620, 0.573, 0.531, and 0.510, respectively. In the analysis using DR genes as the background, the HD gene group contained six significant terms. Five were related to reproduction, and one was "Wnt signaling pathway," which has been implicated in cancer. Our analysis suggests that the HD gene group from PharmGKB is associated with cell communication and drug binding.

No MeSH data available.


Related in: MedlinePlus