GRASP: analysis of genotype-phenotype results from 1390 genome-wide association studies and corresponding open access database

Abstract

Summary: We created a deeply extracted and annotated database of genome-wide association studies (GWAS) results. GRASP v1.0 contains >6.2 million SNP-phenotype association from among 1390 GWAS studies. We re-annotated GWAS results with 16 annotation sources including some rarely compared to GWAS results (e.g. RNAediting sites, lincRNAs, PTMs).

Motivation: To create a high-quality resource to facilitate further use and interpretation of human GWAS results in order to address important scientific questions.

Results: GWAS have grown exponentially, with increases in sample sizes and markers tested, and continuing bias toward European ancestry samples. GRASP contains >100 000 phenotypes, roughly: eQTLs (71.5%), metabolite QTLs (21.2%), methylation QTLs (4.4%) and diseases, biomarkers and other traits (2.8%). cis-eQTLs, meQTLs, mQTLs and MHC region SNPs are highly enriched among significant results. After removing these categories, GRASP still contains a greater proportion of studies and results than comparable GWAS catalogs. Cardiovascular disease and related risk factors pre-dominate remaining GWAS results, followed by immunological, neurological and cancer traits. Significant results in GWAS display a highly gene-centric tendency. Sex chromosome X (OR = 0.18[0.16-0.20]) and Y (OR = 0.003[0.001-0.01]) genes are depleted for GWAS results. Gene length is correlated with GWAS results at nominal significance (P ≤ 0.05) levels. We show this gene-length correlation decays at increasingly more stringent P-value thresholds. Potential pleotropic genes and SNPs enriched for multi-phenotype association in GWAS are identified. However, we note possible population stratification at some of these loci. Finally, via re-annotation we identify compelling functional hypotheses at GWAS loci, in some cases unrealized in studies to date.

Conclusion: Pooling summary-level GWAS results and re-annotating with bioinformatics predictions and molecular features provides a good platform for new insights.

Availability: The GRASP database is available at http://apps.nhlbi.nih.gov/grasp.

Fig. 1.

Proportion of unique SNPs with associations at two thresholds [P ≤ 0.05 (top panel), P ≤ 5E − 8 (bottom panel)] either in, or outside of the MHC class-II region (6p21.3), and/or in, or not in, studies of high phenotype number (eQTL, methylation QTL and metabolomics QTL studies)

Fig. 2.

Pie chart breakdowns of number of results in GRASP by phenotype categories at two statistical threshold (left pies: P ≤ 0.05, right pies: P ≤ 5E − 8). Major categories are cardiovascular traits and risk factors (row A), immunological traits and disorders (row B), neurological traits (row C) and cancer traits (row D)

Fig. 3.

Reported ancestry or ancestries of samples included in GWAS discovery or replication efforts per year from 2002 to 2011

Fig. 4.

GWAS strength of association (blue points) relative to distance to nearest gene for SNPs with P < 1E − 4 and located outside of RefSeq gene transcript isoform boundaries only. eQTL and mQTL and MHC 6p21.3 results are excluded. The inset (upper right) shows enrichment of the proportion of GWAS associations within ∼50 kb of genes (light blue) relative to the proportion of HapMap European imputed SNPs in FHS (yellow)

Fig. 5.

Correlations between gene size and number of GWAS associations for autosomal and sex chromosome SNPs at three statistical thresholds: P ≤ 0.05 (top panel), P ≤ 1E − 4 (middle panel), P ≤ 5E − 8 (bottom panel). Selected gene outliers are labeled for autosomes (black) or sex chromosomes (grey). eQTL and mQTL and MHC 6p21.3 results are excluded

Fig. 6.

UCSC Genome Browser view of prostate cancer associated region at 11q13.3 and annotated lincRNAs and regulatory factors