Untitled Episode

Liang Y et al., The American Journal of Human Genetics, 113 (2026) 276-290. doi:10.1016/j.ajhg.2025.12.014 - TWAS using genetically predicted expression exhibit polygenicity-driven inflation that increases with GWAS sample size and heritability; a gene-specific variance-control correction yields calibrated p values. Key terms: transcriptome-wide association study, TWAS, polygenicity, variance-control, PrediXcan.

Study Highlights:
The authors evaluated TWAS and related xWAS using simulated polygenic null traits and UK Biobank genotypes with predicted mediators including gene expression, metabolites, and brain features. They combined large-scale simulations, theoretical derivations, and empirical regression of mean Z2 on N*h2δ to estimate a gene-specific inflation slope Φ and applied corrections with S‑PrediXcan/PrediXcan across 110 GWAS traits. They show analytically and empirically that Var(Z) ≈ 1 + N*h2δ*Φ, observe a cohort-level slope around 4.2×10^-5, and demonstrate that dividing Z by sqrt(1+N*h2δ*Φ) restores calibration. Applying the variance-control correction yields well-calibrated p values, reduces false positives for highly polygenic traits, and improves precision with minimal loss of power.

Conclusion:
A gene-specific variance-control correction based on an empirically estimated inflation slope Φ corrects polygenicity-driven inflation in TWAS/xWAS and yields calibrated false-positive rates in simulations and real GWAS analyses.

Music:
Enjoy the music based on this article at the end of the episode.

Article title:
A gene-specific variance-control approach corrects polygenicity-driven inflation observed in transcriptome-wide association studies

First author:
Liang Y

Journal:
The American Journal of Human Genetics, 113 (2026) 276-290. doi:10.1016/j.ajhg.2025.12.014

DOI:
10.1016/j.ajhg.2025.12.014

Reference:
Liang Y, Nyasimi F, Im HK. A gene-specific variance-control approach corrects polygenicity-driven inflation observed in transcriptome-wide association studies. The American Journal of Human Genetics. 113 (2026) 276-290. https://doi.org/10.1016/j.ajhg.2025.12.014

License:
This episode is based on an open-access article published under the Creative Commons Attribution 4.0 International License (CC BY 4.0) - https://creativecommons.org/licenses/by/4.0/

Support:
Base by Base – Stripe donations: https://donate.stripe.com/7sY4gz71B2sN3RWac5gEg00

Official website https://basebybase.com

On PaperCast Base by Base you’ll discover the latest in genomics, functional genomics, structural genomics, and proteomics.

Episode link: https://basebybase.com/episodes/twas-variance-control-inflation

QC:
This episode was checked against the original article PDF and publication metadata for the episode release published on 2026-02-06.

QC Scope:
- article metadata and core scientific claims from the narration
- excludes analogies, intro/outro, and music
- transcript coverage: Substantive audit of the transcript’s treatment of TWAS inflation, the Phi inflation parameter, the variance-control correction, and real-data validation across multiple data modalities.
- transcript topics: TWAS/xWAS inflation due to polygenicity; linear scaling of inflation with sample size and trait heritability; gene-specific inflation parameter Phi (Φ); variance-control correction and Z-score calibration (Zcorr = Ztwas / sqrt(1 + Φ N h2δ)); comparison with BACON correction; simulation framework: polygenic null trait; UK Biobank data usage across metabolites and brain features

QC Summary:
- factual score: 10/10
- metadata score: 10/10
- supported core claims: 8
- claims flagged for review: 0
- metadata checks passed: 4
- metadata issues found: 0

Metadata Audited:
- article_doi
- article_title
- article_journal
- license

Factual Items Audited:
- TWAS/xWAS inflation increases with polygenicity
- Inflation scales linearly with GWAS sample size and trait heritability
- Gene-specific inflation parameter Phi (Φ) modulates inflation per gene
- Variance-control method calibrates Z-scores by adjusting with sqrt(1 + Φ N h2δ)
- Variance-control yields calibrated p-values and reduces false positives in polygenic contexts
- Real-data validation across 110 GWAS traits and across metabolites/brain features

QC result: Pass.