Since the beginning of 2020, SARS-CoV-2 infection and its disease, COVID-19, have caused the largest contemporary pandemic to date. While many efforts are being devoted to the characterization of the genetic architecture of COVID-19 effects on human host, its underlying molecular basis has not been exhaustively explored across multiple molecular layers. To understand host response and to prioritize treatment targets, we sought to identify human genes influencing genetically-driven disease risk and severity. To this end, we performed ancestry-aware, trans-layer, multi-omic analyses by integrating recent (April 8, 2022) COVID-19 Host Genetics Initiative GWAS data from six ancestry endpoints - African, Amerindian, South Asian, East Asian, European and meta-ancestry - with functional maps and QTL catalogs.
We explored 91 GWAS hits (P<5e-7), 28% of which were identified in a single ancestry. We analyzed a comprehensive set of >300 cis QTL maps from ~100 biotype sources for colocalization, including disease-relevant biotypes and contexts; blood of COVID-19 patients, large airway epithelium, and lung cell contexts. Across all GWASs, QTL maps and molecular phenotypes, we identified thousands of colocalizations (PP4>0.75) involving >100 genes.
For full description of findings and methodology please see Oliva et al. 2024 (https://www.medrxiv.org/content/10.1101/2024.09.05.24313137v1)
This repository hosts a R shiny app (https://covidgenes.shinyapps.io/shiny/) to interact with the results of the colocalization analysis.
This application was developed by John Lee, in collaboration with, and with contributions of, Meritxell Oliva (colocalization analysis), Reza Hammond (locuszoom plots generation) and Jack Degner (summary statistics display)