arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2603.18256cs.LGcs.AI
收藏

MolRGen:一种用于从头分子生成的训练和评估设置

MolRGen: A Training and Evaluation Setting for De Novo Molecular Generation with Reasonning Models

Université Paris-Saclay(巴黎萨克雷大学);ÉTS Montréal(蒙特利尔ÉTS);ILLS – International Laboratory on Learning Systems(学习系统国际实验室);Mila – Quebec AI Institute(魁北克人工智能研究所);Mistral AI;LIVIA;CNRS, CentraleSupélec(国家科学研究中心,中央超导大学)

Philippe Formont, Maxime Darrin, Ismail Ben Ayed, Pablo Piantanida

AI总结:

本文提出MolRGen,一种用于训练和评估基于推理的分子生成模型的基准测试平台,通过多目标优化提示结合结合亲和力评分和分子性质,评估从头生成的分子,并引入多样性感知的top-k指标和GRPO算法提升性能。

摘要:

近期基于推理的大型语言模型在具有可验证结果的任务上表现出色,但在从头分子生成领域受限于缺乏能够无需参考分子计算奖励的训练环境。我们引入MolRGen,一种用于训练和评估推理LLM的分子验证器,包含约4500个蛋白质口袋靶标,生成约5万个多目标优化提示,结合结合亲和力评分和分子性质如QED、合成可及性、logP和物理化学描述符。不同于基于描述词生成或分子编辑的基准测试,MolRGen在生成时计算分子奖励以评估从头生成的分子。我们基准测试了通用和化学专用的开源LLM,并引入多样性感知的top-k指标以衡量模型是否能生成多样化的高分分子。最后,我们使用验证器微调了一个128B的LLM,展示了改进的性能,但以多样性探索的权衡为代价。MolRGen为研究基于验证器的推理和强化学习在分子设计中的应用提供了可扩展的测试平台。

英文摘要:

Recent reasoning-based large language models have shown strong performance on tasks with verifiable outcomes, but their use in de novo molecular generation remains limited by the lack of training environments where rewards can be computed without reference molecules. We introduce MolRGen, a benchmark and molecular verifier for training and evaluating reasoning LLMs on de novo molecular generation. MolRGen contains approximately 4,500 protein-pocket targets, resulting in 50k multi-objective optimization prompts combining docking scores with molecular properties such as QED, synthetic accessibility, logP, and physicochemical descriptors. Unlike caption-based generation or molecule-editing benchmarks, MolRGen evaluates molecules proposed from scratch by computing rewards at generation time. We benchmark general-purpose and chemistry-specialized open-source LLMs and introduce a diversity-aware top-k metric to measure whether models can generate a diverse set of high-scoring molecules. Finally, we use the verifier to fine-tune a 128B LLM with GRPO, showing improved performance, at the cost of a diversity-exploitation trade-off. MolRGen provides a scalable testbed for studying verifier-based reasoning and reinforcement learning in molecular design.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑