Skip to main content

Genetics of High Cognitive Abilities

Final Report Summary - GHCA (Genetics of High Cognitive Abilities)

Although there are many types of cognitive ability, they correlate substantially. General cognitive ability (g) indexes this covariance and is an important predictor of many key life outcomes. The PI’s 40-year programme of research has contributed to a once controversial finding that is now widely accepted: Individual differences in g are substantially influenced by DNA differences between individuals. Nothing would advance the field more than identifying some of the genes responsible for the heritability of g. Although GWA studies are beginning to identify a few of the genes responsible for genetic influence on g, most of the heritability is ‘missing’, as is the case throughout the life sciences for complex traits and common disorders.

The ERC Advanced Grant, Genetics of High Cognitive Abilities (GHCA, 295366), used two innovations to narrow the missing heritability gap. First, GHCA greatly increased power by studying a large sample of individuals with extremely high g. We have obtained DNA from a unique sample of more than 2000 individuals with an average IQ of 170, which represents the top 99.997th percentile in the population. No other sample of this type is available in the world. Second, GHCA goes beyond reliance on common-SNP microarrays to assess all variants in exomes, including rare variants. GHCA is the first study to examine rare exomic variants on g.

Although we had proposed to sequence exomes, following the launch of GHCA the Illumina Human Exome BeadChip became available, which made it possible to detect exomic variants as rare as .1% much more precisely, efficiently and inexpensively than exome sequencing. We genotyped the GHCA sample of 2000 extremely high-IQ individuals on the Exome BeadChip and obtained a control sample of 6000 individuals with data on g and the same Exome BeadChip. The most exciting result is that the extremely high g sample harbours fewer rare functional variants, which is consistent with the hypothesis that rare functional alleles are more likely to be detrimental than beneficial to g.

Nonetheless, our results suggest that rare exonic variants are not likely to be a major source of missing heritability for g. For this reason, we genotyped the GHCA sample of 2000 extremely high g individuals on the Illumina OmniExpress BeadChip, which provides excellent genome-wide coverage of common SNPs. Our GWA analyses of the extremely high-IQ GHCA sample yielded stronger results than previous GWA analyses of much larger samples of individuals in the normal range of IQ. We also showed that GHCA GWA results add to the polygenic prediction of g independent of other GWA studies.

Our GWA genotyping of both rare exomic variants and genome-wide common variants also provided scaffolding for the first whole-genome sequencing analysis of the 20 highest-IQ individuals in our GHCA sample of extremely high-IQ individuals. The results confirmed our findings from analyses of the Exome Bead Chip that high-IQ individuals have significantly fewer functional mutations as compared to normal-IQ individuals.