Nov 2026· Journal of Chemical Information and Modeling· 0 citations· 108 references
TL;DR
This work developed a compound prioritization strategy that utilizes sampled docked poses from complementary docking approaches into multiple conformations of the target protein to train two multilayer perceptron (MLP) models, and found that a combination of scoring and prediction metrics enriches for hit compounds across library sizes.
Abstract
Compound ranking in structure-based computer-aided drug discovery notoriously yields highly ranked false positives due to sampling and, more frequently, scoring errors. We developed a compound prioritization strategy that utilizes sampled docked poses from complementary docking approaches, i.e., targeted physics-based docking and blind docking with a generative model, into multiple conformations of the target protein to train two multilayer perceptron (MLP) models. The MLPs predict binders or activators at the orthosteric ligand-binding pocket of the nuclear receptor Liver Receptor Homologue-1 (LRH-1, NR5A2). Our approach circumvents the reliance on a single docked pose for scoring compounds or individual scoring metrics for compound ranking. In an out-of-distribution compound set, we observed that the MLPs identified compounds that are chemically dissimilar from those in the training set and are sensitive to single scaffold modifications, making the MLPs potential tools for lead optimization. We applied the MLPs on a pilot study prospective drug discovery campaign, which resulted in the identification of four putative LRH-1 binders. We found that a combination of scoring and prediction metrics enriches for hit compounds across library sizes. In all, this work presents a method to leverage structural and experimental data to aid drug discovery for a challenging protein target.
Predicting how a ligand binds to its target is central to small-molecule drug discovery. Modern cofolding models can generate diverse candidate poses, but reliably ranking these candidates remains challenging. We introduce Pyrite, a pairwise pose-ranking benchmark comprising 1,500 pose pairs across diverse protein-liga...
Megha Srivastava, Michael H. Lim, Allison M. Keys et al.· 1 citation
The results establish that diversity-aware search broadens the set of docking-predicted interaction hypotheses generated under a fixed budget; they do not experimentally establish distinct binding modes or biochemical activity.
Zacharie Bugaud· Journal of Chemical Informat...· 0 citations
DockM8 is, to the authors' knowledge, the most comprehensive open-source platform to integrate the complete consensus structure-based virtual screening pipeline within a single, graphical-interface-driven tool, yielding over 55 million configurable protocols, a breadth unmatched by previously published consensus dockin...
Antoine Lacour, Hamza Agha, Anna K. H. Hirsch et al.· Journal of Cheminformatics· 4 citations
Abstract The computational burden of molecular docking becomes prohibitive when receptor flexibility is explicitly considered, despite the accuracy gained in virtual screening workflows. To mitigate this limitation, we introduce a three-dimensional deep neural network designed to estimate binding energies before exhaus...
D. Ruiz, Renata De Paris, C. V. Quevedo et al.· Journal of Integrative Bioin...· 0 citations
Conventional large-scale structure-based virtual screening typically relies on fixed-receptor docking, in which ligand poses are sampled against a predefined receptor conformation. Protein-ligand co-folding models can instead generate candidate complex conformations by jointly modeling ligand placement and local recept...
Wei-Xin Tian, Xu-Jun Zhang, Odin Zhang et al.· International Journal of Mol...· 0 citations
Structure-based virtual screening needs a receptor conformation that fits the ligand class, but for most targets no ligand-bound crystal is available -- and a single predicted model or an apo structure often misses the binding-competent state. We present ANTIOCH-VS, an open-source framework that runs an entire in silic...
Hale S. Hasdemir, Venkata Sai Sreyas Adury, Pratyush Tiwary University of Maryland Institute for Health Computing et al.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.