Here, we extend the metaGRS strategy to predict IS by incorporating GWAS summary statistics for stroke and its aetiological subtypes along with GWAS summary statistics for risk factors and co-morbidities of IS. This new IS metaGRS is validated and compared with previously published GRS using UK Biobank29,30. We next compare the predictive capacity of the IS metaGRS to established non-genetic risk factors for IS. Finally, we assess the additional information provided by the metaGRS in combination with current guidelines for the treatment of established IS risk factors and create joint models which predict absolute risk of incident IS.
