Biological insights into schizophrenia from ancestrally diverse populations
Article meta
Article Date: 21 January 2026
Article URL: https://www.nature.com/articles/s41586-025-10000-6
Article Image: (not provided)
Summary
This Nature study performs large-scale, multi-ancestry genomic analyses of schizophrenia using datasets from African (AFR), European (EUR) and East Asian (EAS) ancestry cohorts (including MVP, CSP#572, GPC, COGS, PAARTNERS, MGS and PsychAD among others). The work identifies ancestry-specific and shared association signals, improves fine-mapping by leveraging cross-ancestry data, links risk variants to nominated causal genes and brain cell types, and evaluates polygenic risk score (PRS) performance across ancestries. The authors make summary statistics and extensive supplementary data and code available for reproducibility.
Key Points
- Large multi-ancestry GWAS: analyses combine AFR, EUR and EAS datasets to detect both shared and ancestry-specific schizophrenia loci.
- Novel and replicated loci: several genes highlighted with distinct signals across ancestries (examples include PLXNA4, PMAIP1, MACROD2, CD226, DOCK4, GRIN2B, GRIN2A, CLEC16A, SLC14A2, TRPA1).
- Cross-ancestry fine-mapping increases resolution: methods such as SuSiE-R, SuSiEx and MESuSiE were used to prioritise putative causal variants across populations.
- Functional convergence by cell type: TWAS and enhancer–gene linking (ABC-MAX) point to risk genes active in specific brain cell types (excitatory/inhibitory neurons, oligodendrocytes, astrocytes, immune cells).
- SNP heritability and allele-frequency context: SNP-h2 estimates and partitioning show how shared and population-specific variants contribute to observed heritability.
- PRS portability and clinical signals: AFR-trained PRS were tested for association with clinical lab measures (LabWAS) and evaluated across screened controls and cases.
- Extensive data and code availability: summary statistics on Synapse, locus-specific views on LocusZoom, and pipelines using PLINK2, SHAPEIT4, Minimac4, RFMix2, GCTA, METASOFT, MAGMA, scDRS, S-PrediXcan, PRS-CSx and more.
Content summary
The authors assembled diverse cohorts and performed single-ancestry and multi-ancestry GWAS, identifying genome-wide significant associations in AFR-specific analyses and across combined analyses. They used local-ancestry inference and tract-based approaches to dissect signals in admixed individuals and compared directional concordance with previous PGC-EUR results.
Using state-of-the-art fine-mapping and TWAS pipelines, the study prioritised candidate causal genes and connected variants to likely target genes through enhancer–promoter maps. Cell-type specific analyses showed consistent cross-ancestry TWAS correlations in major brain cell types and supported convergent biology in synaptic and neuronal processes. The work also quantified SNP-based heritability within and between ancestries, revealing contributions from shared and ancestry-unique variants.
The authors evaluated PRS performance across ancestries (using PRS-CSx) and explored phenotype associations with electronic health record laboratory measures. All summary statistics, supplementary tables and figures, and code references are provided (Synapse, LocusZoom links and Supplementary Information). Ethical approvals, cohort acknowledgements and funding sources (including the VA Million Veteran Program and multiple NIH grants) are fully documented in the Supplementary Information.
Context and relevance
This paper addresses a major gap in psychiatric genomics: the over-representation of European-ancestry samples. By integrating ancestrally diverse data, the study improves discovery power for loci relevant to under-represented populations, enhances fine-mapping precision, and tests how genetic risk prediction translates across genetic ancestries. Its results are directly relevant to researchers developing causal hypotheses, translational studies aiming to prioritise drug targets, and efforts to make polygenic prediction clinically equitable.
Methodologically, the study demonstrates practical workflows for multi-ancestry GWAS, local-ancestry deconvolution and multi-population fine-mapping — useful templates for future cross-population genetic studies across complex traits.
Author style
Punchy: This is a heavyweight, well-documented multi-ancestry schizophrenia genetics study that shifts the field from EUR-centric discovery to genuinely diverse genomic inference. If you work on psychiatric genetics, functional follow-up, or PRS translation, the full methods and supplementary tables are worth a deep read — the paper both discovers loci and provides the datasets and pipelines to build on.
Why should I read this?
Look — if you care about making schizophrenia genetics relevant beyond European cohorts, this paper saves you weeks of chasing datasets and methods. It shows where risk signals are shared, where they differ, and gives you the tools and links to dig into the loci and cell-type evidence yourself.