Answers are grounded in this paper and cite their sources.
The rational design of covalently binding inhibitors to specifically block selected target proteins has become a common strategy in drug discovery. Such design is based on structural insights into how electrophilic ligands bind to nucleophilic residues. CovPDB v1.0 provided high-resolution cocrystal structures of biologically relevant covalent protein-ligand complexes curated from the Protein Data Bank. Here, we describe CovPDB v2.0, a substantially extended release that addresses three main limitations of the original resource. First, to reflect the numerous novel covalently binding ligands developed and structurally elucidated over the past 4 years, we have classified, annotated, and integrated these entries into the database. This expands the collection from ~2300 to ~6200 protein-ligand complexes, comprising ~3600 pre-reactive ligands, ~1300 target proteins, 119 warheads, and 27 reaction types. Second, we expanded the structural scope beyond experimental cocrystals by generating AlphaFold3 cofolding models for additional ~6800 covalent complexes for which we had experimental evidence for a covalent ligand binding. Third, we added ~7100 non-covalent complexes in which the binding-site cysteines are annotated with a predicted reactivity and ligandability scores. A redesigned web interface allows users to browse, filter, and cross-search experimental, modelled, and ligandable-cysteine entries. CovPDB 2.0 is freely accessible at https://www.pharmbioinf.uni-freiburg.de/covpdb.
No takes yet. Share an insight, caveat, or question.
Feng et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: