AIMNet2(Score) Enables Label-Free Protein-Ligand Binding Scoring

A new preprint introduces AIMNet2(Score), a model that scores protein-ligand binding without experimental data using a single complex structure. Evaluations show it matches or outperforms traditional methods on FEP+ and HiQBind, and generalizes well to 1000-atom kinase sites with minimal error.

2026-07-24 ~ 2026-07-24 · 3 related posts

• AIMNet2(Score) ranks protein–ligand binding from one structure and no affinity labels — olexandr · 2026-07-24
• KIN66 benchmark shows AIMNet2(2025) reaches 3 kcal/mol RMSE on ~1,000-atom kinase sites — olexandr · 2026-07-24
• AIMNet2(Score) matches or beats docking and SQM on FEP+ and HiQBind sets — olexandr · 2026-07-24