Retrieval of binding sites across the AlphaFold human proteome using protein language model representations
Year: 2026
Type: dataset
Abstract: Supporting data for "Retrieval of binding sites across the AlphaFold human proteome using protein language model representations." PocketScope represents every P2Rank-detected cavity in the AlphaFold human proteome as a per-residue tensor built from frozen ESM-C 600M embeddings plus a physicochemical descriptor, and retrieves candidate off-targets by exact late-interaction MaxSim over the whole pocketome. This deposit contains the pocket index, a (153805, 64, 1165) float16 array, the row-to-accession mapping, the benchmark result tables, the evaluation pair sets, and the published figures with their source scripts.
Authors: Mohan, Keshav, Bhargava, Yash
Institutions: Bhargava Systems Research Inc., Johns Hopkins University
Language: English
Subfield: Drug Discovery
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7204684728