A TWAS transcriptional model was computed for each RNA phenotype in each GTEx tissue, and TWAS was performed using FUSION and summary stats for each of 114 GWAS traits previously harmonized to GTEx v8 variant reference. Associations between all RNA phenotypes and all traits in each tissue were filtered with a genome-wide P-value threshold of 5e-8.
Results for all six modalities of RNA phenotypes generated with Pantry, plus residual data-driven phenotypes generated with LaDDR.
Results for xTWAS on full data-driven phenotypes generated with LaDDR.
cis-xQTLs were mapped for each GTEx tissue using tensorQTL and Pantry's cross-modality mapping technique, so they are conditionally independent, even across modalities, within each tissue-gene pair. They are ranked in the order of discovery using stepwise regression per tissue-gene pair, which roughly corresponds to descending association strength. Only the top variant per QTL is shown, which is not necessarily the causal variant.
cis-xQTLs from hybrid mapping of all six modalities of RNA phenotypes generated with Pantry, plus residual data-driven phenotypes generated with LaDDR.
cis-xQTLs for full data-driven phenotypes generated with LaDDR.