Leveraging primary health care data from Estonian Biobank to find novel genetic associations

Genome-wide association studies (GWAS) have significantly advanced the understanding of genetic mechanisms underlying complex human diseases and traits by systematically identifying genetic variants linked to diverse phenotype traits across diverse populations. Large-scale analyses that combine multiple phenotypes are especially valuable, as they can reveal shared genetic architectures and patterns of comorbidity, refining disease classification and risk prediction.

Here, we conducted comprehensive GWAS analyses based on Estonian Biobank EHR data, focusing on 4,884 ICD-10-based disease phenotypes, in a cohort of 206,159 Estonian Biobank participants. By analysing their genotype data, altogether 18,977,777 SNV and indel variants (including common, low-frequency and rare variants), our analyses revealed 2,127 unique genome-wide significant loci, including 778 putatively novel locus-phenotype associations.

Further investigation into coding variants revealed 835 significant associations, including well-established ACMG pathogenic variants and multiple putatively novel associations. Notably, a missense variant in SCN11A , which encodes the Na _v 1.9 sodium channel involved in pain signalling, was associated with decreased migraine risk, while an Estonian-enriched GOT1 variant considerably affected aspartate transaminase enzymatic function.

Our study highlights the potential of population-based biobanks in discovering both common and rare genetic associations, contributing to the identification of novel disease associated loci and expanding the catalog of human disease related genetic variants.