Single-nucleus TWAS maps brain disorder risk across ancestries

September 23, 2026 at 08:26 PM1 min

This digest was compiled by AI from multiple sources — links to the originals are below.

Researchers built single-nucleus transcriptomic imputation models from over 6 million nuclei across 1,494 donors in the PsychAD Consortium. The models were applied to 12 neuropsychiatric and neurodegenerative disorder GWAS to identify cell-type-specific gene–trait associations. Findings were validated in a phenome-wide association study of about 600,000 Million Veteran Program participants.

Key Facts

The PsychAD Consortium atlas includes over 6 million nuclei from 1,494 donors with and without major neuropsychiatric diagnoses.
Researchers trained 94 single-nucleus transcriptomic imputation models across three ancestries and 32 cellular populations in the dorsolateral prefrontal cortex.
The models were applied to 12 neuropsychiatric and neurodegenerative disorder GWAS.
Validation used a phenome-wide association study of approximately 600,000 Million Veteran Program participants.

Single-Nucleus Imputation Models

The PsychAD Consortium generated a population-scale single-nucleus RNA-seq atlas of the dorsolateral prefrontal cortex. The atlas comprises over 6 million nuclei from 1,494 donors spanning European, African, and admixed American ancestries. Using quality-controlled genotype and snRNA-seq data, researchers trained 94 single-nucleus transcriptomic imputation models across three ancestries and 32 cellular populations. For downstream TWAS, analyses were restricted to models with robust cross-validated performance within each ancestry.

Application to Brain Disorders

The single-nucleus models were applied to 12 neuropsychiatric and neurodegenerative disorder GWAS. This single-nucleus TWAS identified cell-type-specific gene–trait associations, including previously unreported disease-linked loci. Findings were validated in a phenome-wide association study of approximately 600,000 participants in the Million Veteran Program. Ancestry-matched models were used to compare effect patterns across populations and characterize pleiotropy of cell-type-specific genetically regulated expression.