Positional mapping approaches were used to identify additional genes that may be involved in MD etiology, including identification of the nearest gene to lead MD variants, aggregating genetic associations across gene regions using fastBAT (see key resources table), and linking associated loci to genes through Hi-C chromatin interactions using Hi-C coupled MAGMA (H-MAGMA) (see key resources table). Furthermore, the gene prioritization method PsyOPS was used to score genes based on prior information on mutational constraint, brain expression, and involvement in neurodevelopmental disorders (Table S4). Of the 18,737 genes assessed using fastBAT, 1,568 were associated with MD (p < 2.67 × 10−6) with the strongest evidence of association at the dopamine receptor D2 (DRD2) gene (p = 9.39 × 10−47). DRD2 was also associated with MD by H-MAGMA in all four brain tissue profiles analyzed (p = 1 × 10−10 to 1 × 10−15). An additional 1,033 genes were also identified as associated with MD based on three-dimensional chromatin data using H-MAGMA. While PsyOPS prioritized a neighboring gene, NCAM1 (PsyOPS score = 0.402), DRD2 had an equivalent score (0.399). Other genes with high PsyOPS prioritization scores were PTPRT, SLC12A5, RFX3, ELAVL2, HCN1, KIF5A, and SHANK3.
