For each trait, we defined the lead IMPACT regulatory annotation as the annotation capturing the greatest per-SNP heritability, for example, the largest, while significant, τ* estimate (Supplementary Table 9). Identifying functional annotations enriched strongly for heritability is an important step to prioritizing regulatory variants for risk prediction models. With the top 5% of SNPs, lead IMPACT annotations captured an average of 43.3% of common SNP heritability (s.e.m. = 2.8%) across these 95 polygenic traits (Extended Data Fig. 3 and Methods). With the top 5% of EUR SNPs, the T-bet TH1 annotation captured 97.1% (s.d. = 17.6%) of asthma heritability. The B cell TBP annotation captured 65.9% (s.d. = 12.1%) of RA heritability. The prostate cancer cell line (LNCAP) TFAP4 annotation captured 60.4% (s.d. = 8.9%) of PrCa heritability. The GATA1 PBmC annotation captured 72.4% (s.d. = 6.0%) of MCV heritability. Lastly, the lung MXI1 annotation captured 31.6% (s.d. = 3.0%) of height heritability; notably, within the MXI1 gene lies a genome-wide significant variant associated with height50. Overall, we observed higher heritability enrichments for autoimmune diseases and hematological traits than for
