Reproduce “ToxBench: A Binding Affinity Prediction Benchmark with AB-FEP-Calculated Labels for Human Estrogen Receptor Alpha” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Meng Liu, Karl Leswing, Simon K. S. Chu, Farhad Ramezanghorbani, Griffin Young, Gabriel Marques, Prerna Das, Anjali Panikar and 14 more
ToxBench provides AB-FEP binding-affinity labels for ERα ligands. DualBind trains on these labels and protein–ligand structures, combining affinity-label fitting with denoising score matching. On the held-out ERα test split, DualBind reaches an RMSE of 2.392 kcal/mol, below Chemprop’s 3.502 and AEV-PLIG’s 2.645 kcal/mol. These errors are against capped AB-FEP labels, not experimental affinities or evidence of transfer beyond ERα.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.