Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2605.28862 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910267224358912 |
|---|---|
| author | Li, Lingxiao Zhang, Haobo Fan, Ruohao Chen, Bin Zhou, Jiayu |
| author_facet | Li, Lingxiao Zhang, Haobo Fan, Ruohao Chen, Bin Zhou, Jiayu |
| contents | Drug discovery is a lengthy and resource-intensive process composed of multiple stages. Among these stages, lead optimization plays a critical role in transforming early hit compounds into viable drug candidates. This stage requires improving ADMET-related properties through subtle structural refinement while preserving key molecular substructures responsible for binding affinity to disease targets. Recent advances in artificial intelligence have shown promise in accelerating various aspects of drug discovery; however, most existing approaches to lead optimization rely on one-step molecular optimization, which fail to account for the long-term consequences of sequential design decisions. To address this limitation, we propose TRACE, a trajectory-aware, LLM-reasoning agent for molecular lead optimization that formulates tool selection as a sequential decision-making problem over action trajectories. Given a lead molecule and an optimization objective, TRACE makes trajectory-aware decisions over molecular optimization tools, enabling forward-looking refinement under structural constraints. Experiments on multiple ADMET optimization tasks show that our agent achieves higher optimization success, larger property improvements, and higher validity, while preserving molecular similarity compared to baseline models. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_28862 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Molecular Lead Optimization via Agentic Tool Planning Li, Lingxiao Zhang, Haobo Fan, Ruohao Chen, Bin Zhou, Jiayu Machine Learning Quantitative Methods Drug discovery is a lengthy and resource-intensive process composed of multiple stages. Among these stages, lead optimization plays a critical role in transforming early hit compounds into viable drug candidates. This stage requires improving ADMET-related properties through subtle structural refinement while preserving key molecular substructures responsible for binding affinity to disease targets. Recent advances in artificial intelligence have shown promise in accelerating various aspects of drug discovery; however, most existing approaches to lead optimization rely on one-step molecular optimization, which fail to account for the long-term consequences of sequential design decisions. To address this limitation, we propose TRACE, a trajectory-aware, LLM-reasoning agent for molecular lead optimization that formulates tool selection as a sequential decision-making problem over action trajectories. Given a lead molecule and an optimization objective, TRACE makes trajectory-aware decisions over molecular optimization tools, enabling forward-looking refinement under structural constraints. Experiments on multiple ADMET optimization tasks show that our agent achieves higher optimization success, larger property improvements, and higher validity, while preserving molecular similarity compared to baseline models. |
| title | Molecular Lead Optimization via Agentic Tool Planning |
| topic | Machine Learning Quantitative Methods |
| url | https://arxiv.org/abs/2605.28862 |