Missing genetic diversity impacts variant prioritisation for rare disorders

Whole genome sequencing identifies millions of genetic variants per individual. When applied to rare disease diagnosis, potentially pathogenic variants are prioritised for clinical interpretation, a process that may be influenced by an individual’s genetic ancestry. We analysed millions of rare protein-altering variants prioritised in 29,425 participants with rare disease from the UK 100,000 Genomes Project. We observed disparities in the number of variants prioritised across genetic ancestry groups, with an up to 3-fold increase in participants with African compared to European ancestries. Variants prioritised in participants with non-European ancestries were less likely to be assessed as pathogenic. Leveraging a cohort of 34,701 diverse genomes from the UK, we identified thousands of candidate variants that were ultra-rare or unobserved across populations in gnomAD but common among ancestry-matched individuals. Our findings highlight the importance of using reference databases that reflect patient genetic diversity when prioritising variants for rare disease diagnosis.