BIG-Graph: Brain Imaging Genetics by Graph Neural Network

Javad Ansarifar, Iliana I. Karipidis, Manish Saggar, David S. Hong

Keywords: Imaging Genetics, Graph Neural Network, Brain Network, genome-wide association studies

Abstract: Imaging genetics is one of the foremost emerging fields in neuroscience research that aims to combine neuroimaging and genetic information with phenotypes to shed light on inherent underlying mechanisms. While significant progress has been made in integrating brain imaging, like functional magnetic resonance imaging (fMRI), with genetic data, such as single nucleotide polymorphisms (SNPs), little progress has been made in studying them jointly using graph structures. To raise a new perspective and overcome challenges in analyzing data with high dimensionality and inherently complex relationships, we developed a graphical neural network model (BIG-Graph) that jointly learns to effectively represent both neuroimaging and genetic data in a nonlinear manner without any prior knowledge. Here, we demonstrate that joint learning of imaging-genetics using BIG-Graph largely outperforms existing state-of-the-art Imaging genetics models and networks trained separately on neuroimaging or genetic data in predicting a variety of phenotypes.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

Submission Guidelines: Yes

Please Choose The Closest Area That Your Submission Falls Into: Neuroscience and Cognitive Science (e.g., neural coding, brain-computer interfaces)