How well does whole genome sequencing improve ability to detect association with asthma in candidate genes compared to existing GWAS platforms in African American populations?
Rafaels NM, Johnston HR, Huang L, Chavan S, Wilson JG, Williams KL, Ware LB, Ober C, Meyers DA, Hartert TV, Foreman M, Ford JG, Gonza Burchard E, Bleecker ER, Taub M, Beaty TH, Ruczinski I, Mathias RA, Barnes KC. How well does whole genome sequencing improve ability to detect association with asthma in candidate genes compared to existing GWAS platforms in African American populations?. J Allergy Clin Immunol 2015; 135(2 Suppl):AB164.
J Allergy Clin Immunol
Rationale: To date, 225 genes from 32 genome-wide association study (GWAS) manuscripts have shown association with asthma, yet only 2 GWAS manuscripts were from African ancestry populations. As part of the Consortium on Asthma among African-ancestry Populations in the Americas (CAAPA), 328 African Americans were whole genome sequenced (WGS). Because current GWAS arrays are inadequate for African populations, we propose analyzing candidate genes in CAAPA will help detect new variants associated with asthma.
Methods: Candidate genes were selected searching for asthma in GWAS catalog and phegeni databases with p-value<1E-5. Logistic regression adjusting for population stratification for common single nucleotide polymorphisms (SNPs) and Fisher’s exact test for rare SNPs was performed in asthma candidate genes on 168 asthmatic and 160 non-asthmatic African Americans on WGS samples sequenced with Illumina HiSeq 2000 and GWAS samples genotyped with Illumina OMNI 2.5. Haplotype tagging coverage was determined using Haploview tagger restricting to common SNPs (MAF≥0.01) with r2≥0.8, forcing inclusion of any GWAS SNPs.
Results: Analysis was performed on 80 candidate genes analyzed for 231,548 and 17,586 SNPs from WGS and GWAS panels, respectively. The OMNI panel captured 15,098 of 37,739 tag SNPs. WGS analysis uncovered associations with asthma not tagged by OMNI panel in IL13 (rs2243200; P=3.7E-6; MAF=0.04), SCG3 (rs35776517; P=2.3E-4; MAF=0.24), PRKG1 (rs4576762; P=2.7E-4; MAF=0.05), and C11orf71 (rs181308697; P=4.8E-4; MAF=0.03).
Conclusions: In a sample of 328 African Americans in 80 candidate genes, we demonstrate that a standard GWAS panel captures only 40% of common variation. Association testing with asthma in additional populations of African descent is ongoing.