Sep 2026· Journal of Chemical Information and Modeling· Vol 66 19, pp.
12948-12954
· 0 citations· 24 references
Computer ScienceMedicine
TL;DR
This work systematically benchmarked several molecular docking tools representing all-atom models, highlighting the complementarity of AI-based approaches and methods based on physical sampling in all-atom models, in terms of their applicability range, and argues for the benefits of tighter integration.
Abstract
Generalizability of molecular docking predictions across novel protein targets and distinct small-molecule chemotypes is key to the successful application of deep learning co-folding approaches in drug discovery, as highlighted by recent benchmarks, including Runs-n-Poses [Škrinjar, P.; et al.Nat. Struct. Mol. Biol.2026, 33, 782-794.]. To compare generalizability across physics-based, hybrid rescoring models, and co-folding approaches, we systematically benchmarked several molecular docking tools representing these paradigms. Our results support previous observations that co-folding approaches show superior performance on targets resembling their training datasets. However, they deteriorate sharply to unacceptably low 20-40% success rates on novel dissimilar systems, consistent with memorization. In contrast, physics-based and hybrid methods exhibit greater out-of-distribution robustness, maintaining more than 60% success rates for docking protein-ligand complexes with minimal similarity to previously known complexes. Our results highlight the complementarity of AI-based approaches and methods based on physical sampling in all-atom models, in terms of their applicability range, and argue for the benefits of tighter integration.
Boltz is benchmarked using a curated set of ligand-bound human G protein-coupled receptors from families unseen during training, showing that while Boltz generally predicts receptor backbones accurately, ligand poses can contain significant errors that lead to a limited ability to reproduce experimental affinity data w...
Lichirui Zhang, R. Friesner, Edward B. Miller et al.· npj Drug Discovery· 0 citations
It is shown that prospective structure selection, rather than structure generation, represents the primary bottleneck in ensemble-based VS, highlighting an urgent need for novel structural descriptors to identify high-performing conformations.
Jaeoh Shin, K. Joo, Jejoong Yoo· Journal of Chemical Informat...· 1 citation
KinConfBench is introduced, a curated benchmark of 2225 high-quality human kinase chains to evaluate the ability of four state-of-the-art cofolding models—Boltz-2, Chai-1, Protenix, and RoseTTAFold-All-Atom—to recover both canonical and rare conformational states.
Kun-Yang Sun, T. Head-Gordon· npj Drug Discovery· 1 citation
Novo-1, a coarse-grained cofolding framework for binding- affinity prediction, offers more than one order of magnitude speed-up over the leading open-source baseline, Boltz-2, and demonstrates meaningful selectivity, separating the binding affinities of identical compounds between on-targets and related off-targets.
Nikhil Shenoy, David Errington, Emmanuel Bengio et al.· bioRxiv· 3 citations· ⚡1
These findings demonstrate that robust peptide-binder prediction requires careful consideration of training-set redundancy, target conformational flexibility, and complementary confidence metrics when selecting computational methods for de novo candidate prioritization.
Mark Fonteyne, Nada Badr, P. Kuppen et al.· Digital Discovery· 0 citations
Related blog posts
MIT News · Artificial Intelligence· news.mit.eduNov 9, 2026
A new machine-learning framework aims to improve the success rate of computational protein design while moving away from results that reproduce sequences found in nature.