Optimal Molecular Design: Generative Active Learning Combining REINVENT with Absolute Binding Free Energy Simulations

Hannes Loeffler,Shunzhou Wan,Marco Klähn,Agastya Bhati,Peter Coveney

DOI: https://doi.org/10.26434/chemrxiv-2024-sr1v6

2024-04-26

Abstract:Active learning (AL) is a specific instance of sequential experimental design and uses machine learning to intelligently choose the next data point or batch of molecular structures to be evaluated. In this sense it closely mimics the iterative design-make-test-analysis cycle of laboratory experiments to find optimized compounds for a given design task. Here we describe an AL protocol which combines generative molecular AI, using REINVENT, and physics-based absolute binding free energy molecular dynamics simulation, using ESMACS, to discover new ligands for two different target proteins, 3CLpro and TNKS2. We have deployed our generative active learning (GAL) protocol on Frontier, the world’s only exa-scale machine. We show that the protocol can find better binders compared to baseline, a surrogate ML docking model for 3CLpro and compounds with experimentally determined binding affinities for TNKS2. The ligands found are also chemically diverse and occupy a different chemical space than the baseline. We vary the batch sizes that are put forward for free energy assessment in each GAL cycle to assess the impact on their efficiency on the GAL protocol and recommend their optimal values in different scenarios. Overall, we demonstrate a powerful capability of the combination of physics-based and AI methods which yields effective chemical space sampling at an unprecedented scale, of immediate and direct relevance for modern, data-driven drug discovery.

Chemistry

What problem does this paper attempt to address?

The paper aims to address the issue of molecular design optimization in drug discovery, specifically by accelerating the process of finding high-affinity ligand molecules for specific protein targets through the combination of generative artificial intelligence and physics-based absolute binding free energy simulations. Specifically, the research team developed an active learning (AL) protocol that combines REINVENT generative molecular AI with ESMACS (a method for absolute binding free energy calculations based on molecular dynamics simulations) to identify new ligands for two different protein targets—3CLpro (the main protease of coronaviruses) and TNKS2 (an enzyme associated with various cellular processes). By deploying this generative active learning (GAL) protocol on the world's only exascale computer, Frontier, the researchers demonstrated that the method could find better binding molecules than the baseline model, and these molecules exhibit chemical diversity, occupying different chemical spaces from the baseline. Additionally, they explored the impact of different batch sizes on the efficiency of the GAL protocol and recommended optimal values for different scenarios. In summary, the study demonstrates the ability to effectively sample chemical space on an unprecedented scale by combining physics-based methods with AI technology, which has direct and significant implications for modern data-driven drug discovery.

Optimal Molecular Design: Generative Active Learning Combining REINVENT with Absolute Binding Free Energy Simulations

Optimal Molecular Design: Generative Active Learning Combining REINVENT with Precise Binding Free Energy Ranking Simulations

Human-in-the-loop active learning for goal-oriented molecule generation

Active Learning Guided Drug Design Lead Optimization Based on Relative Binding Free Energy Modeling

Sample Efficient Reinforcement Learning with Active Learning for Molecular Design

MF-LAL: Drug Compound Generation Using Multi-Fidelity Latent Space Active Learning

Effective drug-target affinity prediction via generative active learning

ChemSpaceAL: An Efficient Active Learning Methodology Applied to Protein-Specific Molecular Generation

Learning Binding Affinities via Fine-tuning of Protein and Ligand Language Models

Optimizing Drug Design by Merging Generative AI With Active Learning Frameworks

Fragment-Based Ligand Generation Guided By Geometric Deep Learning On Protein-Ligand Structure

Optimization of binding affinities in chemical space with generative pre-trained transformer and deep reinforcement learning

Generative AI for Drug Discovery: A GPT-2 and LSTM Based Models for Designing EGFR Inhibitors

Automated design of multi-target ligands by generative deep learning

Guided Multi-objective Generative AI to Enhance Structure-based Drug Design

Generative Active Learning for the Search of Small-molecule Protein Binders

Reinforced Genetic Algorithm for Structure-based Drug Design

Efficient Exploration of Chemical Space with Docking and Deep Learning

Modern hit-finding with structure-guided de novo design: identification of novel nanomolar A2A receptor ligands using reinforcement learning

Binding Affinity Prediction with 3D Machine Learning: Training Data and Challenging External Testing