Population-level genetic variation shapes generative brain mechanisms

The structural organisation of the human brain emerges probabilistically as we develop. Due to the inherent complexity of the human brain, understanding the forces that shape this probabilistic emergence remains one of the central challenges of systems theory and neuroscience. Across 2153 children (9-11 years old) we used a computational model to simulate the formation of structural brain connectivity, conceptualised as a trade-off between the cost of new connections {eta} and their topological value {gamma}. We then triangulated this population-level neuroimaging and computational modelling with genomics. For each participant we assessed their genetic propensity for cognitive ability by calculating polygenic scores. Modelled parameters differed systematically for participants depending upon their genetic propensity. Those with the highest genetic propensity had a significantly weaker {eta} term - put simply, their networks emerged with a weaker distance penalty. Strikingly, this softer distance penalty produces more stochastic, diverse, and efficient networks. Furthermore, across the sample, overlapping biological and cellular pathways between polygenic scores and each childs optimal {eta}-{gamma} trade-off emerged. This application of computational modelling demonstrates a converging genomic basis for structural brain development and cognitive ability across the population, providing a mechanistic explanation of how and why characteristic network topologies emerge from children at the extreme distributions of polygenic scores, and why they might predict cognitive ability.