The prediction accuracies of polygenic scores in humans vary depending on the characteristics of the samples, as well as based on the study design, within a single ancestry group.

A resource to systematically explore the association between overall genetic risk for 162 risk factors and disease with 551 outcomes.

Correction for stratification in genome-wide association studies should be informed by the demographic history of the sample.

Current polygenic scores for education do not provide accurate predictions of educational achievement above other data or at the individual level.

Across urate, IGF-1, and testosterone, thousands of dispersed variants contribute to heritability, while previously characterized core pathways are highly enriched for associations and reveal feedback mechanisms governing serum biomarker levels.

Polygenic selection signals in humans estimated from previously existing GWAS should be viewed with caution due to concerns about residual population stratification.

Many prior signals of polygenic adaptation on human height do not replicate in the UK Biobank.

Admixture dynamics have recently changed over time in Mexican Americans resulting in an impact on the genetic architecture of phenotypic variation.

Genomic associations with lifespan principally reflect heart disease/smoking/dementia but not other cancers, and distinguish lifespan differences of five years between top/bottom deciles of a score derived from DNA alone.

Genetic data revealed existing drugs that could be repurposed to improve lung function, with evidence that compounds which control blood glucose could be particularly useful for respiratory impairment.