Leveraging genomic and transcriptomic data of diverse ancestry to uncover mechanisms of psychiatric risk in the adult and developing brain

To better understand molecular mechanisms underlying psychiatric disorders, we need improved analytic approaches for integrating large-scale genomic data with brain-based transcriptomics. However, one critical component of individual variation - different levels of gene regulation due to genetic ancestry diversity - has not been traditionally incorporated into such analyses. To address this, we leveraged the ancestral diversity of individuals in adult and developing brain Genotype-Expression (GEx) reference panels of the PsychENCODE Project and psychiatric GWAS of the Psychiatric Genomics Consortium to enhance detection of GReX (Genetically Regulated gene expression) and transcriptome-wide association study (TWAS) signals. To investigate alternative GEx-level choices, we trained GReX models using rigorously constructed subsets of the human postmortem dorsolateral prefrontal cortex GEx panel, generated through downsampling, segregating, and mixing samples of Admixed African (AA) and European (EUR) ancestries, while considering disease status in the subset design. TWAS-based gene trait associations (GTAs) were obtained by integrating these GReX models with ancestry-specific GWASs of bipolar disorder (BIP), major depressive disorder (MDD), post-traumatic stress disorder (PTSD) and schizophrenia (SCZ). Genes with ancestry-specific GReX were enriched in specialized pathways involving mitochondrial functions, organelle structure, and metabolism, while genes with GReX in both ancestries demonstrated high concordance (>95%) in predictor SNP weight directions. Applying GReX from either ancestry to a single ancestry GWAS produced GTAs with concordant effects sizes while each uncovering unique FDR significant trait associated signals at the gene and pathway levels. GTAs based on AA GWAS meta analyzed with GTAs based on EUR GWAS enhanced signals and alleviated noise. EUR-specific GReX produced TWAS pathways that included corticosteroid signaling in PTSD, TGF-beta and neurotrophins in MDD, and inflammation and viral life cycle in SCZ. AA-specific GReX produced TWAS pathways that included glutamine signaling in PTSD, proline-peptide DNA activity in MDD, and immune cytotoxicity, serotoninergic and dopaminergic pathways in SCZ. Genes with ancestry-specific GReX in adult and developing brain were part of similar pathways. Finally, the developing brain TWAS were enriched in specialized developmental and neuronal pathways and produced a higher proportion of shared signals between ancestries. In conclusion, we demonstrate the benefits of leveraging diverse ancestral backgrounds in TWAS analysis, provide insights into which genes and pathways are better captured by ancestry-specific panels, and advocate for genomic region-specific TWAS integration strategies over a uniform genome-wide approach to uncover molecular mechanisms.