Toward Closed-loop Molecular Discovery via Language Model, Property Alignment and Strategic Search

2025-12-10

AI Summary

Drug discovery faces challenges including low efficiency, poor success rates, and difficulty in multi-objective optimization. This paper introduces Trio, a novel language model–driven fragment-based molecular generation framework. Trio integrates context-aware fragment assembly, policy-gradient reinforcement learning, and Monte Carlo Tree Search (MCTS) to jointly optimize multiple properties—binding affinity, drug-likeness, and synthetic accessibility—under protein pocket constraints. Its core innovation lies in modeling molecular design as an interpretable, strategy-guided search process, overcoming the limitations of single-objective optimization. Experiments demonstrate that all generated molecules are 100% chemically valid. Compared to state-of-the-art methods, Trio achieves improvements of 7.85% in binding affinity, 11.10% in drug-likeness, and 12.05% in synthetic accessibility, while expanding molecular diversity by over fourfold.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Constraint Satisfaction and Optimization: Constraint Learning and Acquisition
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Search and Retrieval-Augmented AI: Web search models and ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Drug discovery is a time-consuming and expensive process, with traditional high-throughput and docking-based virtual screening hampered by low success rates and limited scalability. Recent advances in generative modelling, including autoregressive, diffusion, and flow-based approaches, have enabled de novo ligand design beyond the limits of enumerative screening. Yet these models often suffer from inadequate generalization, limited interpretability, and an overemphasis on binding affinity at the expense of key pharmacological properties, thereby restricting their translational utility. Here we present Trio, a molecular generation framework integrating fragment-based molecular language modeling, reinforcement learning, and Monte Carlo tree search, for effective and interpretable closed-loop targeted molecular design. Through the three key components, Trio enables context-aware fragment assembly, enforces physicochemical and synthetic feasibility, and guides a balanced search between the exploration of novel chemotypes and the exploitation of promising intermediates within protein binding pockets. Experimental results show that Trio reliably achieves chemically valid and pharmacologically enhanced ligands, outperforming state-of-the-art approaches with improved binding affinity (+7.85%), drug-likeness (+11.10%) and synthetic accessibility (+12.05%), while expanding molecular diversity more than fourfold.

Problem

• Overcoming low success rates in drug discovery via generative models
• Addressing poor generalization and interpretability in molecular design
• Balancing binding affinity with pharmacological properties for better ligands

Innovation

• Integrates fragment-based language modeling, reinforcement learning, and Monte Carlo search
• Enables context-aware fragment assembly with physicochemical feasibility
• Guides balanced exploration and exploitation in protein binding pockets