HT Gwern. The important bottom-line is that with a sample of 147k participants, we can now predict 6.9% of phenotypic intelligence. Relevant quotes below, with my emphasis:
This study had four goals: firstly, to facilitate the discovery of new genetic loci associated with intelligence; secondly, to add to our understanding of the biology of intelligence differences; thirdly, to examine whether combining genetically correlated traits in this way produces results consistent with the primary phenotype of intelligence; and, finally, to test how well this new meta-analytic data sample on intelligence predict phenotypic intelligence variance in an independent sample. We apply Multi-Trait Analysis of Genome-wide association studies to three large GWAS: Sniekers et al (2017) on intelligence, Okbay et al. (2016) on Educational attainment, and Hill et al. (2016) on household income. By combining these three samples our functional sample size increased from 78 308 participants to 147 194. We found 107 independent loci associated with intelligence, implicating 233 genes, using both SNP-based and gene-based GWAS.
The effects of age, sex, and population stratification (7 components) were controlled for using residuals extracted from a regression model.
Using our meta-analytic data set on intelligence we carried out polygenic prediction into Generation Scotland: Scottish Family Health Study and found that 6.9% of phenotypic intelligence could be predicted (Table 2), an improvement of 43.75% on previous estimates of 4.8%
HT Gwern. The important bottom-line is that with a sample of 147k participants, we can now predict 6.9% of phenotypic intelligence. Relevant quotes below, with my emphasis: