Model-designed mutation paths switched WW-domain binding specificity
Source check: Source checked. Not independent scientific confirmation.
Finding: The first public preprint appeared on bioRxiv on 2025-12-10. Its reviewed version reports that designed WW-domain mutational paths largely retained binding, while the same mutations introduced in scrambled order disrupted function. The eLife reviewed-preprint v1 (2026-04-09) later quantified binding for 54 of 58 tested designs and described cross-reactive intermediates on some specificity transitions. In-vitro binding to representative peptides does not establish in-vivo function or generalize to all WW domains.
Technology: Unsupervised restricted Boltzmann machine
Publication: Preprint · Test context: direct test. Catalog categories, not an evidence grade.
- AI contribution
- The learned sequence model sampled intermediate sequences and mutation orders between natural WW domains with different binding specificities.
- Test and result
- The primary version history identifies a bioRxiv preprint posted 2025-12-10 before the eLife reviewed-preprint v1 dated 2026-04-09. The eLife v1 reports that 54 of 58 tested sequences designed along model-sampled WW-domain paths bound at least one representative peptide class; variants with the same mutations in scrambled order were nonfunctional. The specific 54/58 count is attributed to the eLife v1, not assumed for the earlier version.
- Limitation / still unknown
- In-vitro binding to representative peptides does not establish in-vivo function or generalize to all WW domains.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
Evidence sources: https://www.biorxiv.org/content/10.64898/2025.12.08.693000v1 · https://elifesciences.org/reviewed-preprints/110491v1
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 22 Nov 2026. Review scope: Read the eLife reviewed-preprint v1 Methods and Results and version history; checked the version history identifying the earlier 2025-12-10 bioRxiv preprint and author affiliations on the eLife article information page..
No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.
Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 3/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.