bioRxivJClub
Computational Design and Evaluation of De Novo Protein Minibinders Targeting Specific Sites of STAT3
bioRxiv · · Preprint · Open access
Sragovicz, Tobias
Abstract ↗AI summary
The abstract is read at the publisher; the summary is JClub's.
Eleven high-confidence STAT3 minibinder designs were generated against seven SH2/linker/DNA-binding/coiled-coil hotspots, and all are predicted by AlphaFold3 to dock to intended domains.
- Why it matters: Small-molecule STAT3 inhibitors have had a high failure rate in clinical trials, motivating alternative binding-based approaches to target STAT3 sites.
- What they did: Using BindCraft, 139 designs passed filters and 11 emerged from five hotspots, then AlphaFold3 predicted docking to targeted hotspot residues, often within 4.5 [A].
- The result: BLASTp found no matches for six binders and only very low-significance matches for five, and some pairs are predicted to dock simultaneously without overlap, enabling potential STAT3 sandwich assays.
The findingWhy it mattersWhat they didThe result
- Open access