AF3Score: A Score-Only Adaptation of AlphaFold3 for Biomolecular Structure Evaluation
Authors:
- 16 July 2025
- journal article
- Published by American Chemical Society (ACS) in Journal of Chemical Information and Modeling
Abstract
Scoring biomolecular complexes remains central to structural modeling efforts. Recent studies suggest that AlphaFold (AF) - a revolutionary deep learning model for biomolecular structure prediction - has implicitly learned an approximate biophysical energy function. While many researchers highly rely on AF-derived scores for structure evaluation, existing AlphaFold2-based implementations require iterative refinement of the input structure, leading to biased scoring. To address this limitation, we adapted AlphaFold3 into a score-only model, AF3Score, by directly feeding input coordinates into the confidence head while bypassing the diffusion-based structure module. AF3Score demonstrates robust performance in structural quality assessment across diverse systems, including monomeric proteins, protein-protein complexes, de novo designed binders, fold-switching proteins, and protein-ligand complexes. In benchmarking designed binder screening, AF3Score outperformed state-of-the-art methods for 8 out of 10 targets. Moreover, combining AF3Score with AlphaFold2-derived methods significantly improved the enrichment of experimentally validated binders, increasing the success rate from 15.2 to 31.6%. Additionally, AF3Score effectively identified stable conformations in fold-switching proteins, whereas AlphaFold predominantly predicted only the dominant fold. These findings highlight the broad applicability of AF3Score, from high-throughput screening in de novo binder design to filtering docking-generated poses and molecular dynamics (MD) trajectories.Keywords
All Related Versions
Funding Information
- Changping Laboratory
This publication has 36 references indexed in Scilit:
- AlphaFold2 fails to predict protein fold switchingProtein Science, 2022
- Energy‐based graph convolutional networks for scoring protein docking modelsProteins: Structure, Function, and Bioinformatics, 2020
- The Rosetta All-Atom Energy Function for Macromolecular Modeling and DesignJournal of Chemical Theory and Computation, 2017
- Simultaneous Optimization of Biomolecular Energy Functions on Features from Small Molecules and MacromoleculesJournal of Chemical Theory and Computation, 2016
- Evolution, energy landscapes and the paradoxes of protein foldingBiochimie, 2015
- Updates to the Integrated Protein–Protein Interaction Benchmarks: Docking Benchmark Version 5 and Affinity Benchmark Version 2Journal of Molecular Biology, 2015
- Learning To Fold Proteins Using Energy Landscape TheoryIsrael Journal of Chemistry, 2014
- Sampling and scoring: A marriage made in heavenProteins: Structure, Function, and Bioinformatics, 2013
- One Target—Two Different Binding Modes: Structural Insights into Gevokizumab and Canakinumab Interactions to Interleukin-1βJournal of Molecular Biology, 2013
- Crystal structure and kinetic analysis of β-lactamase inhibitor protein-II in complex with TEM-1 β-lactamaseNature Structural & Molecular Biology, 2001