Co-expression-based models improve eQTL predictions for transcriptome-wide association studies and highlight new schizophrenia-associated genes - PubMed
3 hours ago
- #co-expression
- #eQTL
- #schizophrenia
- Co-expression models INGENE and MODULE incorporate trans-acting variants within gene networks to improve gene expression prediction.
- Integrating these models with cis-based predictors enhanced imputation accuracy for 18,744 genes across brain regions.
- Application to schizophrenia genotypes identified 766 associated genes, with 641 being new discoveries from transcriptome-wide analyses.
- Findings emphasize the role of distal regulatory effects and gene network interactions in schizophrenia risk.