A data-driven method to study brain structural connectivities via joint analysis of microarray data and dMRI data

Connective structure is an outstanding feature of the brain. Exploring the mechanism of its development and its variation across individual and species might hold the key to understanding brain functions and brain diseases. Genes are widely recognized as the fundamental regulators of the connective architecture, but it is still far from fully understanding genetics impacts on brain connectivities. Quantitative analysis of advanced imaging data, such as diffusion MRI (dMRI), provides phenotypic features and plausible clues for exploring the genetic reasons. Therefore, in this paper, we jointly analyzed dMRI data and microarray gene expression data. By developing a novel method to compare the dMRI derived structural connectivity matrix and gene expression distance matrix, we identified gene groups which might contribute to structural wiring diagram. Effectiveness and reproducibility of this method has been demonstrated.