Genetic underpinnings of brain structural connectome for young adults

Yize Zhao,Changgee Chang,Jingwen Zhang,Zhengwu Zhang

DOI: https://doi.org/10.48550/arXiv.2212.00967

2022-12-02

Abstract:With distinct advantages in power over behavioral phenotypes, brain imaging traits have become emerging endophenotypes to dissect molecular contributions to behaviors and neuropsychiatric illnesses. Among different imaging features, brain structural connectivity (i.e., structural connectome), which summarizes the anatomical connections between different brain regions, is one of the most cutting-edge while under-investigated traits; and the genetic influence on the structural connectome variation remains highly elusive. Relying on a landmark imaging genetics study for young adults, we develop a biologically plausible brain network response shrinkage model to comprehensively characterize the relationship between high dimensional genetic variants and the structural connectome phenotype. Under a unified Bayesian framework, we accommodate the topology of brain network and biological architecture within the genome; and eventually establish a mechanistic mapping between genetic biomarkers and the associated brain sub-network units. An efficient expectation-maximization algorithm is developed to estimate the model and ensure computing feasibility. In the application to the Human Connectome Project Young Adult (HCP-YA) data, we establish the genetic underpinnings which are highly interpretable under functional annotation and brain tissue eQTL analysis, for the brain white matter tracts connecting the hippocampus and two cerebral hemispheres. We also show the superiority of our method in extensive simulations.

Applications