A haplotype reference panel constructed from 490,319 UK Biobank genomes improves genotype imputation for global populations.

A haplotype reference panel built from 490,319 UK Biobank genomes improves genotype imputation accuracy for global populations, especially East and South Asians, with lower phasing switch errors than TOPMed r3.

• Why it matters: Existing reference panels can limit phasing and imputation performance for diverse ancestries and rare variants, constraining downstream association discovery and fine-mapping.
• What they did: Whole-genome sequencing from 490,319 UK Biobank participants was used to create a panel enabling standalone and cloud phasing and imputation via EagleImp 2 and EagleImp-RAP, benchmarked against TOPMed r3 and 1000 Genomes subpopulations.
• The result: The UKB panel lowers phasing switch error rates versus TOPMed r3 for all global superpopulations except AMR and improves MAE for 4 of 5 superpopulations and 12 of 26 subpopulations (MAF down to 0.01%), while reimputation of COVID-19 GWAS datasets shows common-variant imputation remains unsaturated.

AI summary of the abstract · may contain errors · not medical advice · read the paper.