Answers are grounded in this paper and cite their sources.
Absolute binding free energy (ABFE) calculations can evaluate chemically diverse ligands, but their complex workflows remain challenging for nonexpert users and prone to errors. Here, we present ezABFE, an easy-to-use Python workflow for automated ABFE calculations within the AMBER ecosystem. ezABFE minimizes manual intervention in system preparation, representative-structure selection, Boresch restraint atom selection, simulation setup, free-energy estimation, and graphical analysis. The workflow employs both MBAR and TI for alchemical free-energy estimation and additionally provides MM-GBSA and optional MM-PBSA end-point calculations. Benchmarking across T4 lysozyme L99A, BRD4(1), CDK8, and the theophylline-binding RNA aptamer shows good overall agreement with experimentally derived binding free energies and comparable performance of MBAR and TI. Additional tests demonstrate that appropriate Boresch restraint atom selection is important for simulation stability and binding free-energy accuracy, and that ligand partial-charge models and the HMR/4 fs simulation protocol can influence calculated affinities, although the limited comparisons do not support general conclusions regarding either factor. The computationally inexpensive end-point methods also retain practical value for ligand ranking. Overall, ezABFE provides an accessible, automated workflow for reproducible ABFE calculations with improved throughput across diverse receptor systems, including nucleic acid receptors.
No takes yet. Share an insight, caveat, or question.
Li et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: