30-Second Takeaway
- Regional nonsense-constraint substantially sharpens loss-of-function interpretation and Mendelian gene discovery.
- Mechanistic IRF2BPL and STR studies stress non-SNV mechanisms and pleiotropy in neurogenetic disorders.
- Biobank-scale rare-variant and GWAS work support context-specific PRS, survival phenotypes, and drug target nomination.
Week ending February 28, 2026
New genomic tools and studies refining variant interpretation, risk prediction, and neurodegenerative diagnostics
NMD-informed regional nonsense constraint improves interpretation of truncating variants and gene discovery
An NMD-informed regional nonsense constraint metric was derived from sequencing data of 730,947 individuals. Investigators identified 2,764 genes with significant regional nonsense constraint, including 641 autosomal dominant disease genes. In 32,260 rare-disease trios, de novo truncating variants were 9.5-fold enriched in constrained versus unconstrained regions. Such variants in constrained regions conveyed up to 5.9-fold higher odds of achieving a genetic diagnosis. Regional constraint also highlighted 22 candidate disease genes with clustered de novo variants, supporting novel gene discovery and variant prioritization.
IRF2BPL NEDAMSS variants disrupt phase separation and deplete nuclear protein
Pathogenic truncating and missense IRF2BPL variants cluster in a central region containing three low-complexity regions and a higher-complexity domain. The polyA/polyQ-rich LCR1, together with an upstream zinc finger, normally drives liquid-liquid phase separation of IRF2BPL in cells. Disease-associated variants disrupt physiological phase separation, forming aberrant cytoplasmic condensates that sequester wild-type IRF2BPL from the nucleus. Nuclear depletion associates with WNT1 upregulation and altered neuronal electrophysiology, establishing a shared mechanistic pathway for NEDAMSS. These findings support pathogenicity of central-region truncating and select missense variants via a defined biophysical mechanism.
GATE-STAAR enables calibrated rare-variant time-to-event analysis in large WGS biobanks
GATE-STAAR is a frailty model framework for rare-variant association testing of censored time-to-event phenotypes using whole-genome sequencing data. It explicitly handles heavy censoring, cryptic relatedness, and population structure while integrating functional annotations for power and interpretability. Simulations demonstrated proper type I error control with good power in large biobank settings. Applied to approximately 400,000 UK Biobank participants, it uncovered rare-variant associations with age at diagnosis for multiple diseases, replicated in All of Us. This method supports survival-style analyses of rare variants for disease onset and progression in clinical genetics consortia.
Systematic STR screening in dementia yields a 2% diagnostic rate and underscores pleiotropy
Whole-genome sequencing from 1,559 neurodegenerative dementia patients and 1,522 controls was screened for 22 known pathogenic short tandem repeats. After quality control and RP-PCR validation, 33 pathogenic expansions in nine genes were confirmed, accounting for 2.12% of dementia cases. Pathogenic expansions were significantly associated with dementia status, with an odds ratio of 3.57. ExpansionHunter Denovo improved detection of C9orf72 expansions missed by targeted calling, highlighting technical limitations of single tools. Intermediate-length TBP alleles were nominally enriched in progressive supranuclear palsy, suggesting possible contribution from sub-pathogenic repeats.
References
Numbered in order of appearance. Click any reference to view details.
Additional Reads
Optional additional studies from this edition.