Skip to main content
. Author manuscript; available in PMC: 2019 Mar 12.
Published in final edited form as: Science. 2018 Dec 14;362(6420):eaat8464. doi: 10.1126/science.aat8464

Fig. 6. GRNs assign genes to GWAS loci for psychiatric disorders.

Fig. 6.

(A) A schematic depicting how SCZ GWAS loci were assigned to putative genes. The number of SCZ GWAS loci and their putative target genes (SCZ genes) annotated by each assignment strategy is indicated (top). The overlap between SCZ genes defined by QTL associations (QTL), chromatin interactions (Hi-C), and activity relationships (activity) is depicted in a Venn diagram (bottom). SCZ genes with more than two evidence sources were defined as high-confidence (high conf.) genes. (B) A GRN of TFs, enhancers, and 321 SCZ high-confidence genes, on the basis of TF activity linkages. A subnetwork for CACNA1C is highlighted on the right. (C) An example of the evidence indicating that GWAS SNPs that overlap with CHRNA2 eQTLs also have chromatin interactions and activity correlations with the same gene. Orange dots refer to SNPs that overlap between eQTLs and GWAS plots. (D) TFs that are significantly enriched in enhancers (left) and promoters (right) of SCZ genes. FDR, false discovery rate. (E) SCZ genes show higher expression levels in neurons (particularly excitatory neurons) than in other cell types. (F) Brain disorder GWAS show stronger heritability enrichment in brain regulatory variants (eQTLs) and elements (enhancers) than non–brain disorder GWAS. ADHD, attention-deficit/hyperactivity disorder; T2D, type 2 diabetes; CAD, coronary artery disease; IBD, inflammatory bowel disease.