Integrating network analyses and genetics with large-scale RNA-sequencing of schizophrenia brains. M. Fromer1,2 for the CommonMind Consortium, Swedish Schizophrenia Consortium, Schizophrenia Working Group of PGC 1) Genetics and Genomic Sciences, Mount Sinai School of Medicine, New York, NY; 2) Psychiatric Genomics, Mount Sinai School of Medicine, New York, NY.

   The most recent schizophrenia GWAS reported >100 associated loci, implying a high degree of polygenicity. To better understand the pathology of neuropsychiatric disease, we formed the CommonMind Consortium ( to generate large-scale data (RNA-seq, ChIP-seq, DNA-seq/genotyping) from human post-mortem brain samples. Here, we identify functional changes in gene expression using RNA-seq of 554 samples (265 schizophrenia cases and 289 controls) from the dorsolateral prefrontal cortex (BA9/46). Clinical (gender, age of death, medications) and technical (brain bank, post-mortem interval, RNA quality, sequencing batch) covariates, as well as hidden confounders, were controlled using surrogate variable analysis (SVA). Analyses are ongoing, but initial application of linear models implemented in voom/limma identified ~15% of expressed genes as differential between cases and controls (FDR 5%). These genes were nominally enriched for DNA variants associated with schizophrenia, including rare (frequency < 0.1%) nonsynonymous variants in Swedish case-control exome sequencing (p=0.012) and common GWAS loci (p=0.045). De novo loss-of-function (nonsense, frameshift, essential splice site) mutations in autism, intellectual disability, and epilepsy affected the differential genes (p=0.0053, 0.00058, 0.018), though this did not hold for schizophrenia de novos (p=0.12). Preliminary gene coexpression networks constructed using WGCNA (Weighted Gene Co-expression Network Analysis) identified ~40 modules, 11 differentially expressed (FDR 5%). A differential module of ~1000 genes involved in synaptic transmission was seemingly enriched in rare nonsynonymous variants in case-control exomes (p=0.013), and a differential module of ~200 postsynaptic genes related to mitochondrial energy production showed enrichment of common GWAS loci (p=0.0093). The synaptic transmission module also tended to be enriched for loss-of-function mutations in autism, intellectual disability, and epilepsy (p=0.0015, 0.06, 0.035). Genes impacted by loss-of-function mutations in schizophrenia were enriched (p=0.00014), as were common GWAS loci (p=0.0044), in a non-differential glutamatergic signaling module. This large dataset will be made public in early 2015 and will include a catalogue of brain-expressed genes and isoforms, as well as eQTL, from cases and controls. This resource will facilitate novel discoveries relating neurobiology to disease risk and advance therapies.

You may contact the first author (during and after the meeting) at