FRAGMENTA: End-to-end Fragmentation-based Generative Model with Agentic Tuning for Drug Lead Optimization
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914171197587456 |
|---|---|
| author | Suzuki, Yuto Awolade, Paul LaBarbera, Daniel V. Banaei-Kashani, Farnoush |
| author_facet | Suzuki, Yuto Awolade, Paul LaBarbera, Daniel V. Banaei-Kashani, Farnoush |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_20510 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | FRAGMENTA: End-to-end Fragmentation-based Generative Model with Agentic Tuning for Drug Lead Optimization Suzuki, Yuto Awolade, Paul LaBarbera, Daniel V. Banaei-Kashani, Farnoush Artificial Intelligence 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. |
| title | FRAGMENTA: End-to-end Fragmentation-based Generative Model with Agentic Tuning for Drug Lead Optimization |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2511.20510 |