FRAGMENTA: End-to-end Fragmentation-based Generative Model with Agentic Tuning for Drug Lead Optimization

TL;DR

FRAGMENTA introduces an end-to-end, agentic AI framework for drug lead optimization that improves molecule generation diversity and automates tuning through conversational feedback, outperforming traditional methods.

Contribution

It presents a novel fragmentation and generation approach combined with agentic AI for autonomous, efficient drug lead optimization.

Findings

Nearly twice as many high-scoring molecules identified in experiments

Autonomous agentic tuning outperformed human-in-the-loop methods

Effective in real-world cancer drug discovery scenarios

Abstract

Molecule generation using generative AI is vital for drug discovery, yet class-specific datasets often contain fewer than 100 training examples. While fragment-based models handle limited data better than atom-based approaches, existing heuristic fragmentation limits diversity and misses key fragments. Additionally, model tuning typically requires slow, indirect collaboration between medicinal chemists and AI engineers. We introduce FRAGMENTA, an end-to-end framework for drug lead optimization comprising: 1) a novel generative model that reframes fragmentation as a "vocabulary selection" problem, using dynamic Q-learning to jointly optimize fragmentation and generation; and 2) an agentic AI system that refines objectives via conversational feedback from domain experts. This system removes the AI engineer from the loop and progressively learns domain knowledge to eventually automate tuning. In real-world cancer drug discovery experiments, FRAGMENTA's Human-Agent configuration identified nearly twice as many high-scoring molecules as baselines. Furthermore, the fully autonomous Agent-Agent system outperformed traditional Human-Human tuning, demonstrating the efficacy of agentic tuning in capturing expert intent.