To assess enrichment of overlapping genes, we performed a one-sided Kolmogorov-Smirnov (KS) test of the P values for the gene signatures versus all genes and assessed significance via resampling. Significant enrichment was observed (Supplementary Table 2A) for the haloperidol signatures, but not the clozapine signature (P = 1 × 10−9 and 0.29, respectively). We also tested whether the direction of effect for drug signature genes was more concordant than expected by chance; this was tested using a hypergeometric test whose null hypothesis assumes that up- and down-regulated drug-mediated genes were randomly sampled from genes either up- or down-regulated (at any P value) in the CMC SCZ differential expression tests. We found the haloperidol signature significantly less concordant than expected by chance (35 out of 237 genes concordant, one-sided P < 1 × 10−4), while the clozapine signature showed significant concordance, despite the lack of enrichment in the previous test (22 of 31 genes concordant, one-sided P = 0.013). For haloperidol, the enrichment of significance but depletion of concordance is perhaps consistent with a scenario in which such antipsychotic drugs target
