DualBind: A Dual-Loss Framework for Protein-Ligand Binding Affinity Prediction
This work addresses a crucial challenge in drug development by providing a more robust method for predicting binding affinities, though it appears incremental as it builds on existing machine learning approaches.
The paper tackles the problem of protein-ligand binding affinity prediction by introducing DualBind, a framework that combines supervised and unsupervised learning to improve accuracy and reduce reliance on labeled data, achieving enhanced performance in experiments.
Accurate prediction of protein-ligand binding affinities is crucial for drug development. Recent advances in machine learning show promising results on this task. However, these methods typically rely heavily on labeled data, which can be scarce or unreliable, or they rely on assumptions like Boltzmann-distributed data that may not hold true in practice. Here, we present DualBind, a novel framework that integrates supervised mean squared error (MSE) with unsupervised denoising score matching (DSM) to accurately learn the binding energy function. DualBind not only addresses the limitations of DSM-only models by providing more accurate absolute affinity predictions but also improves generalizability and reduces reliance on labeled data compared to MSE-only models. Our experimental results demonstrate that DualBind excels in predicting binding affinities and can effectively utilize both labeled and unlabeled data to enhance performance.