Integrating rare variant genetics and brain transcriptome data implicates novel schizophrenia putative risk genes

Abstract

The etiology of schizophrenia is elusive, in part due to its polygenic nature. Genome-wide association studies (GWAS) have successfully identified hundreds of schizophrenia risk loci, that are pinpointed to over one hundred genes through fine mapping. Besides common variants with relatively small effect size from GWAS, rare variants or ultra rare variants also play a significant role in conferring the schizophrenia risk from SCHEMA (Schizophrenia Exome Sequencing Meta-Analysis) results. However, burden results from SCHEMA study indicate that more new risk genes remain hidden and to be discovered. To boost the power of identifying new risk genes, we integrated genetics from SCHEMA and transcriptome data from BrainSpan using a multi-omics integration tool, DAWN, through which we have identified 47 schizophrenia putative risk genes that include 19 new risk genes, in addition to nearly all SCHEMA risk genes with FDR < 5 %. GO functional enrichment reveals that 47 SCZ putative risk genes are significantly enriched in cell to cell signaling, cell communications, transporter, in line with the hypothesis of two hit schizophrenia model. SynGO analysis suggests 47 schizophrenia putative risk genes are enriched in pre-synapse, synapse and post-synapse, supporting the well established link between synapses and schizophrenia.