Abstract
Background Genetic associations link hematopoietic traits and disease end-points, but most causal variants and genes underlying these relationships are unknown. Here, we used genetic colocalization to nominate loci and genes related to shared genetic signal for hematopoietic, cardiovascular, autoimmune, neuropsychiatric and cancer phenotypes.
Results Our findings recapitulate developmental hematopoietic lineage relationships, identify loci associating traits with causal genetic relationships, and reveal novel associations. Out of 2706 loci with genome-wide significant signal for at least 1 blood trait, we identified 1779 unique sites (66%) with shared genetic signal for 2+ hematologic traits at a false discovery rate <5%. We could assign some sites to specific developmental cell types during hematopoiesis based on affected traits, including those likely to impact hematopoietic progenitor cells and/or megakaryocyte-erythroid progenitor cells. Through an expanded analysis of 70 human traits, we define 2+ colocalizing traits at 2007 loci from an analysis of 9852 sites (20%) containing genome-wide significant signal for at least 1 GWAS trait. In addition to variants and genes underlying shared genetic signal between blood traits and disease phenotypes that had been previously related through mendelian randomization studies, we define loci and related genes underlying shared signal between eosinophil count and eczema. We also identified colocalizing signals in a number of clinically relevant coding mutations, including in sites linking PTPN22 with Crohns disease, NIPA with coronary artery disease and platelet trait variation, and the hemochromatosis gene HFE with altered lipid levels. Finally, we anticipate potential off-target effects on blood traits related novel therapeutic targets, including TRAIL.
Conclusions Our findings provide a road map for gene validation experiments and novel therapeutics related to hematopoietic development, and offer a rationale for pleiotropic interactions between hematopoietic loci and disease end-points.
