# Generalization Beyond Benchmarks: Evaluating Learnable Protein-Ligand Scoring Functions on Unseen Targets

arXiv: https://arxiv.org/abs/2512.05386 · HTML: https://mlsanity.com/paper/2512.05386/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 8 of its 57 references that are in the map.
- First posted on arXiv 5 Dec 2025.

## Builds on (most cited first, 8 of 8)

- [Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges](https://mlsanity.com/paper/2104.13478/) (2021)
- [DiffDock: Diffusion Steps, Twists, and Turns for Molecular Docking](https://mlsanity.com/paper/2210.01776/) (2022)
- [ChemBERTa-2: Towards Chemical Foundation Models](https://mlsanity.com/paper/2209.01712/) (2022)
- [Development and evaluation of a deep learning model for protein-ligand binding affinity prediction](https://mlsanity.com/paper/1712.07042/) (2017)
- [Ankh: Optimized Protein Language Model Unlocks General-Purpose Modelling](https://mlsanity.com/paper/2301.06568/) (2023)
- [Deep Confident Steps to New Pockets: Strategies for Docking Generalization](https://mlsanity.com/paper/2402.18396/) (2024)
- [DeepAtom: A Framework for Protein-Ligand Binding Affinity Prediction](https://mlsanity.com/paper/1912.00318/) (2019)
- [Data Leakage and Redundancy in the LIT-PCBA Benchmark](https://mlsanity.com/paper/2507.21404/) (2025)

## Built on by (most cited first, 1 of 1)

- [Out-of-Distribution Detection in Molecular Complexes via Diffusion Models for Irregular Graphs](https://mlsanity.com/paper/2512.18454/) (2025)

## Abstract (arXiv)

As machine learning becomes increasingly central to molecular design, it is vital to ensure the reliability of learnable protein-ligand scoring functions on novel protein targets. While many scoring functions perform well on standard benchmarks, their ability to generalize beyond training data remains a significant challenge. In this work, we evaluate the generalization capability of state-of-the-art scoring functions on dataset splits that simulate evaluation on targets with a limited number of known structures and experimental affinity measurements. Our analysis reveals that the commonly used benchmarks do not reflect the true challenge of generalizing to novel targets. We also investigate whether large-scale self-supervised pretraining can bridge this generalization gap and we provide preliminary evidence of its potential. Furthermore, we probe the efficacy of simple methods that leverage limited test-target data to improve scoring function performance. Our findings underscore the need for more rigorous evaluation protocols and offer practical guidance for designing scoring functions with predictive power extending to novel protein targets.