Multi-ancestry fine-mapping of the chromosome 17q12-q21 asthma locus identifies independent associations implicating lymphocyte and eosinophil levels in the causal pathway

Background Asthma pathophysiology varies by age-of-onset and involves diverse immune processes reflected in white blood cell (WBC) subsets. To investigate the genetic architecture of asthma and potential endophenotypes, we analyzed the chr17q12-q21 locus, a robustly replicated asthma locus, across European (EUR), African (AFR), East Asian (EAS), and South Asian (SAS) ancestry groups from the UK Biobank (UKB) and Biobank Japan (BBJ). The largest EUR sample was further stratified by age-ofonset as a proxy for etiological heterogeneity. Results Eight independent asthma signals were identified in UKB-EUR, including two novel associations (Signal 2—rs72832915 and Signal 6—rs507671). Signal 4, corresponding to the originally identified pediatric signal, showed the strongest cross-ancestry evidence, with asthma risk and diminished lymphocyte count co-localizing in three populations. Signal 8 was distinguished by multiple lines of evidence converging on rs112401631 as a likely causal variant, including fine-mapping, colocalization with eosinophil and lymphocyte counts, and mediation of asthma risk through eosinophil count. Signal 6 implicated RARA expression, suggesting vitamin A metabolism impacts on late-onset asthma, the only associated stratum. Notably, integrating WBC traits and Bayesian fine-mapping enabled leveraging non-European ancestry groups to strengthen causal inference despite smaller sample sizes. Conclusion These findings illustrate how combining age-of-onset stratification, quantitative endophenotypes, and multi-ancestry analyses can reveal mechanistic heterogeneity and prioritize specific variants and path-ways for functional validation. This framework is broadly applicable to complex diseases with measurable quantitative endophenotypes.