RiboZero Gold kit [Illumina]) was followed as for the primary human CMC cohort. Sequencing data were processed similarly as for the human CMC cohort, with reads aligned to the macaque reference genome and transcriptome (mmul1), but with two minor changes: STAR77 was used for efficient alignment, and featureCounts78 was used for gene-level quantification. rRNA rates were all below 1%. RNA expression levels were normalized using voom, and limma differential expression analysis was performed, adjusting for sex and RNA isolation batch, to assess the effects of haloperidol treatment (N = 17 in total, grouped to increase statistical power) and clozapine (N = 9) drug groups, as compared to the baseline untreated group (N = 8). While no genes were considered differentially expressed after multiple test correction, we used a nominal P ≤ 0.01 cutoff to identify signatures for haloperidol and clozapine treatment, which resulted in human-orthologous gene sets of size 237 and 31, respectively.
