用图神经网络预测化学反应机理，准确率达99%且可解释。
DeepMech: A Machine Learning Framework for Chemical Reaction Mechanism Prediction
- 基于原子与键级注意力机制，结合反应模板引导生成机理。
- 在3万条反应数据上，小步步骤预测准确率98.98%，完整机理达95.94%。
- 可识别关键反应位点，适合用于药物设计和原始生命化学研究。
化学反应机理（CRM）的完整步骤预测仍是重大挑战。传统方法依赖专家实验或高成本量子计算，而现有深度学习模型常忽略关键中间体和步骤，易产生幻觉。我们提出DeepMech，一种基于图结构的可解释深度学习框架，采用原子与键级注意力机制，并受广义反应操作模板（TMOps）引导，以生成完整的反应机理。该模型在自建的ReactMech数据集（约3万条含10万条原子映射且质量守恒的基元步骤）上训练，基元步骤预测准确率达98.98±0.12%，完整机理预测准确率为95.94±0.21%。即使在分布外场景及副产物/侧产物预测中仍保持高保真度。扩展至前生物化学中的多步机理，成功重建了从简单前体到丝氨酸和醛戊糖的两条路径。注意力分析结果与化学直觉一致，揭示了关键反应原子/键，使模型具备可解释性，适用于反应设计。
原文摘要 · Abstract (English)
Prediction of complete step-by-step chemical reaction mechanisms (CRMs) remains a major challenge. Whereas the traditional approaches in CRM tasks rely on expert-driven experiments or costly quantum chemical computations, contemporary deep learning (DL) alternatives ignore key intermediates and mechanistic steps and often suffer from hallucinations. We present DeepMech, an interpretable graph-based DL framework employing atom- and bond-level attention, guided by generalized templates of mechanistic operations (TMOps), to generate CRMs. Trained on our curated ReactMech dataset (~30K CRMs with 100K atom-mapped and mass-balanced elementary steps), DeepMech achieves 98.98+/-0.12% accuracy in predicting elementary steps and 95.94+/-0.21% in complete CRM tasks, besides maintaining high fidelity even in out-of-distribution scenarios as well as in predicting side and/or byproducts. Extension to multistep CRMs relevant to prebiotic chemistry, demonstrates the ability of DeepMech in effectively reconstructing 2 pathways from simple primordial substrates to complex biomolecules such as serine and aldopentose. Attention analysis identifies reactive atoms/bonds in line with chemical intuition, rendering our model interpretable and suitable for reaction design.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。