MFBind: a Multi-Fidelity Approach for Evaluating Drug Compounds in Practical Generative Modeling

TL;DR

MFBind is a multi-fidelity framework that combines fast docking and accurate free energy calculations to improve drug compound evaluation in generative modeling, balancing accuracy and computational efficiency.

Contribution

It introduces a multi-fidelity deep surrogate model with active learning, integrating docking and binding free energy methods for better drug discovery predictions.

Findings

MFBind outperforms existing surrogate models in accuracy.

It significantly improves the quality of generated drug compounds.

The approach reduces computational costs while maintaining high prediction fidelity.

Abstract

Current generative models for drug discovery primarily use molecular docking to evaluate the quality of generated compounds. However, such models are often not useful in practice because even compounds with high docking scores do not consistently show experimental activity. More accurate methods for activity prediction exist, such as molecular dynamics based binding free energy calculations, but they are too computationally expensive to use in a generative model. We propose a multi-fidelity approach, Multi-Fidelity Bind (MFBind), to achieve the optimal trade-off between accuracy and computational cost. MFBind integrates docking and binding free energy simulators to train a multi-fidelity deep surrogate model with active learning. Our deep surrogate model utilizes a pretraining technique and linear prediction heads to efficiently fit small amounts of high-fidelity data. We perform extensive experiments and show that MFBind (1) outperforms other state-of-the-art single and multi-fidelity baselines in surrogate modeling, and (2) boosts the performance of generative models with markedly higher quality compounds.