Analysis of Gene Sets Extracted from a GWAS of the UK Biobank

The GWAS analysis of the UK Biobank by the Neale Lab provided SNPs associated with over 4,000 unique phenotypes. With these GWAS summary statistics we were able to create gene sets using a nearest-gene method. Utilizing gene-gene correlation data along with previously published information about phenotype-gene pairs, we found that a p-value cutoff of 5e-8 and no distance cutoff was optimal for creating gene sets from the UK Biobank GWAS. We were able to create gene sets for 1,120 phenotypes with a total unique gene coverage of 14,003 and an average of 12.5 genes per phenotype. We then used Enrichr to perform gene set enrichment analysis for each phenotype. To publish these results, we developed a Twitter bot called EnrichrBot to automatically interact with the SbotGWA, a Twitter bot created by the Neale Lab to broadcast their UK Biobank GWAS analysis. EnrichrBot listens to SbotGWA and tweets the enrichment analysis results for the genes extracted from the GWAS analysis posted by SbotGWAS for a specific phenotype.
This presentation is by Allison Seiden, an undergraduate student at Johns Hopkins University. Allison describes her summer research project with the BD2K-LINCS DCIC in the Ma'ayan Lab at the Icahn School of Medicine at Mount Sinai.
lincs-dcic.org/...
icahn.mssm.edu/...
/ botenrichr
/ sbotgwa
github.com/Maa...