187-LB: Leveraging Information across Multiple Related Glycaemic Traits Improves Fine-Mapping Resolution Free

Glycaemic traits fasting insulin (FI), fasting glucose (FG), 2-hour glucose (2hGlu) and glycated haemoglobin (HbA1c) are used to diagnose and monitor type 2 diabetes (T2D). MAGIC (Meta-Analysis of Glucose and Insulin-related traits Consortium) identified 242 loci associated with these traits in 281,416 individuals, from 6 ancestries. European (EUR) ancestry fine-mapping resulted in credible sets accounting for 99% of the posterior probability of association (99CS) with a median of 55 SNPs. Trans-ancestry (TA) fine-mapping led to 99CS with a median of 45 SNPs. Here, we investigated whether fine-mapping resolution could be improved by leveraging data across multiple related traits. We used the novel method FLexible And SHared information Fine-Mapping (flashfm), which takes advantage of sharing of information between traits. We compared its performance with single-trait FINEMAP results. We conducted EUR fine-mapping for 53 loci associated with ≥2 glycaemic traits (p<10^-6, N=119 loci-trait associations). EUR ancestry FINEMAP results were used as input for flashfm. Default parameters were used, except a prior odds of 80% (TOdds=0.25) of a shared causal variant across traits was used. 99CS were used to compare single-trait and flashfm results. Flashfm successfully fine-mapped 48 loci (106 locus-trait associations, 5 loci could not be run). At 40/106 locus-trait associations flashfm led to 99CS with ≤10 SNPs, of which 4 had a single SNP, prioritising SNPs for functional follow-up. Fine-mapping across ≥2 traits reduced the number of SNPs in the 99CS in 80.1% (85/106) of locus-trait associations. The median number of SNPs in the 99CS were significantly reduced from 50 to 19 (p=1.1 x10^-13). In 47.3% (44/93) of locus-trait associations flashfm reduced the 99CS size compared to TA fine-mapping. Our results show that leveraging data across multiple related traits is a useful approach to improve single-ancestry fine-mapping resolution, and that it may also improve TA fine-mapping.