BRAIN-MAGNET: A functional genomics atlas for interpretation of non-coding variants

Genome-wide assessment of genetic variation is becoming routine in genetics, yet functional interpretation of non-coding single nucleotide variants in both common and rare diseases remains a major challenge. Here, we used chromatin immunoprecipitation coupled to self-transcribing active regulatory region sequencing (ChIP-STARR-seq) to functionally annotate non-coding regulatory elements (NCREs) in cellular models of human brain development. This provides gene regulatory insights into neural stem cells and evidence of NCRE priming already in embryonic stem cells for later neural activity. Based on this functional genomics atlas, we developed BRAIN-MAGNET (brain-focused artificial intelligence method to analyze genomes for non-coding regulatory element mutation targets), a functionally validated convolutional neural network that predicts NCRE activity from DNA sequence composition and identifies nucleotides required for NCRE function. BRAIN-MAGNET allows fine-mapping of genome-wide association study (GWAS) loci for common neurological traits and prioritizing candidate disease-causing rare non-coding variants in genetically unexplained individuals with neurogenetic disorders. Together, this NCRE atlas and BRAIN-MAGNET represent a powerful resource for the interpretation of non-coding genetic variation, possibly aiding the identification of previously unrecognized enhanceropathies.