Many thanks to @gregor-research.bsky.social @simonsfoundation.org @uthealthsph.bsky.social

Many thanks to our co-first authors: Jiaoyang Xu and Jacob Schmidt

🧰 Resources
We also provide a new interactive NMD-escape protein feature viewer to support variant interpretation:
👉 github.com/schmidtjacob...

🧬 Why This Matters
Our study shows:
➡️ Many pathogenic truncating variants do not undergo NMD, producing stable, altered proteins.
➡️ These effects may arise through structural remodeling and phase separation changes, not simple haploinsufficiency.

5️⃣ Variant-level modeling supports altered phase-separation propensity.
Using biophysical modeling, we demonstrate that NMD-escape variants in genes like KAT6B and RARB dramatically change predicted phase separation propensity.

4. We identified 25 NMD-escape genes that are also haploinsufficient neurodevelopmental disorder (NDD) genes with 28% of them encoding condensate-forming proteins, highlighting a potential link between NMD-escape and biomolecular condensate dysfunction.

3. Pathogenic NMD-escape variants :
Truncate/alter deeper into functional domains
Reshape amino-acid composition (↑ glycine/proline, altered aromatics)
Alter intrinsic disorder potential consistent with changes in interaction multivalency and disorder-driven signaling.

2️⃣ NMD-escape regions in those genes are enriched for functional “hotspots.”
These regions contain significantly more: Protein–protein interaction interfaces, Low-complexity sequences, and regulatory elements (PFAM domains, NLS motifs, PTMs)
than matched control genes.

1️⃣ We identified 148 autosomal-dominant disease genes enriched for predicted NMD-escape variants. These include nonsense, −1 frameshift, and +1 frameshift alleles pointing to NMD-escape as a widespread pathogenic mechanism, not a rare exception.