bioRxivJClub
A structural screen for candidate polyphosphate binding surfaces in human membrane associated proteins
bioRxiv · · Preprint
Wu, Jia
Abstract ↗AI summary
The abstract is read at the publisher; the summary is JClub's.
A computational workflow nominated 82 human membrane-associated protein entries as candidate noncovalent polyphosphate binders from 19,481 docking searches across 5,648 AlphaFold models.
- Why it matters: Polyphosphate can modulate protein function via noncovalent binding, but structure-based binding partners among human membrane-associated proteins lack systematic, searchable identification methods.
- What they did: The workflow screens predicted structures for phosphate-chain accommodating surfaces using custom scores integrating surface electrostatics, docking results, and structural confidence across 2,925 UniProt entries.
- The result: Predicted polyP contacts in 18 of the 82 candidates overlapped experimentally characterized functional residues, enabling targeted experimental testing of specific binding interfaces.
The findingWhy it mattersWhat they didThe result