Recent advances in genomics are producing powerful DNA predictors of complex traits, such as cognitive abilities, educational outcomes or mental health. In this study, we created genetic predictors, so-called polygenic scores, of intelligence and educational achievement (GCSE scores), using current state-of-the-art approaches. 

Polygenic scores aggregate the information of thousands of genetic variants across the genome into a score representing the genetic predisposition an individual carries for a particular trait.

We showed that we are now able to detect ~10% and ~15% of the variability in intelligence and educational achievement in the population by using measured DNA variants alone. For intelligence this corresponded to a difference of 15 IQ points between individuals with very high values vs. very low values (top and bottom deciles) of the polygenic score, on average. While for educational achievement, this corresponded to an average ‘C’ grade for the lowest decile and an average ‘A’ grade for the highest decile. We also showed that there is no difference between girls and boys in their genetic propensity to both intelligence and educational achievement.

It is important to keep in mind, however, that polygenic scores speak more about general propensities within populations, rather than individuals' destiny. For example, there is considerable overlap in outcomes between individuals in the top and bottom polygenic score deciles. That is, individuals with a low genetic score for educational achievement can achieve an A* grade in their GCSE exams and individuals with high genetic score get ‘C’ grade or below for the GCSE exams. The same applies for intelligence. 

Nevertheless, we showed that polygenic scores for educational attainment and intelligence are currently the most powerful predictors in the behavioural sciences, even more predictive than family background. These polygenic scores are useful tools for examining gene-environment interplay in educational outcomes, and this is part of our future research program. 

To read the full article, click here: https://doi.org/10.1038/s41380-019-0394-4