Retrieval of binding sites across the AlphaFold human proteome using protein language model representations
- 1Bhargava Systems Research Inc.
Abstract
Protein language models (PLMs) provide powerful representations of protein sequence, but their utility for proteome-scale binding-site retrieval remains unclear. Here, we present PocketScope, a training-free framework that represents cavity-lining residues using frozen ESM-C 600M embeddings and retrieves related binding sites through exhaustive late-interaction MaxSim, without pooling or approximate nearest-neighbor search. PocketScope identified 153,805 cavities across 37,682 proteins in the AlphaFold human proteome and recovered documented drug off-targets across a curated set of pharmacological pairs. On the ProSPECCTs benchmark, PocketScope ranks 1st of 23 methods by mean rank across the ten collections. PocketScope provides a practical framework for proteome-scale off-target prediction. PocketScope is open source and also freely available as a web server at https://www.bhargavaresearch.org/pocketscope.
Related
- Proteome scale drug prioritization from AlphaFold 3 interaction confidence
- A tale of two states: AlphaFold binding pockets on the apo-holo continuum
- RNA-binding proteome dynamics following ischemia-reperfusion injury in the kidney
- The Neurophysiology of Binding and Retrieval in Action Control and Its Alterations in Tourette Syndrome
- CHORUS: A proteome-scale atlas of membrane interaction interfaces constructed using cross-protein language model embeddings from sequences alone
BibTeX
@misc{mohan2026retrieval,
title = {{Retrieval of binding sites across the AlphaFold human proteome using protein language model representations}},
author = {K. Mohan and Y. Bhargava},
howpublished = {bioRxiv},
year = {2026},
month = aug,
doi = {10.64898/2026.08.30.748136},
url = {https://www.biorxiv.org/content/10.64898/2026.08.30.748136v1},
note = {CC BY-NC-ND},
}
RIS
TY - UNPB
TI - Retrieval of binding sites across the AlphaFold human proteome using protein language model representations
AU - K. Mohan
AU - Y. Bhargava
PY - 2026
DA - 2026/08/30
PB - bioRxiv
DO - 10.64898/2026.08.30.748136
UR - https://www.biorxiv.org/content/10.64898/2026.08.30.748136v1
AB - Protein language models (PLMs) provide powerful representations of protein sequence, but their utility for proteome-scale binding-site retrieval remains unclear. Here, we present PocketScope, a training-free framework that represents cavity-lining residues using frozen ESM-C 600M embeddings and retrieves related binding sites through exhaustive late-interaction MaxSim, without pooling or approximate nearest-neighbor search. PocketScope identified 153,805 cavities across 37,682 proteins in the AlphaFold human proteome and recovered documented drug off-targets across a curated set of pharmacological pairs. On the ProSPECCTs benchmark, PocketScope ranks 1st of 23 methods by mean rank across the ten collections. PocketScope provides a practical framework for proteome-scale off-target prediction. PocketScope is open source and also freely available as a web server at https://www.bhargavaresearch.org/pocketscope.
ER -