Investigating the effects of genetic risk of schizophrenia on behavioural traits.

To verify the reliability of these polygenic associations, we repeated the analyses using PRS comprising only the 108 sentinel genome-wide significant schizophrenia variants 15 , based on the assumption that the effect size estimates of genome-wide significant variants should be less affected by population structure than non-significant variants; 77 of the 104 associated traits were nominally significant ( P < 0.05) based on the PRS comprising only genome-wide significant variants.

also in 7,414 other papers

We also tested associations with several ‘control traits’, such as ‘leg pain on walking’, ‘breast fed’ and ‘month attended baseline assessment centre’ (see Methods) for comparison, since epidemiological testing on such large sample sizes are vulnerable to producing highly significant associations that are the consequence of small biases or statistical artefacts.

also in 127,188 other papers