用磁性超图建模分子相互作用，提升药物毒性预测准确率
ChemHyperMag: Physics-informed magnetic hypergraph learning improves molecular ADMET prediction
- 构建基于官能团的有向超图，捕捉分子非对称互动
- 在少标签情况下超越现有方法，无需构象采样
- 可解释的磁相位信号揭示分子方向性特征
准确预测药物的吸收、分布、代谢、排泄和毒性（ADMET）对新药研发至关重要。现有方法多采用无向分子图与成对边，忽略了非对称相互作用、不可逆动态及官能团与环系的基元效应。本文提出ChemHyperMag，用于在标签缺失情况下的多任务ADMET预测。该模型基于环、BRICS片段、Bemis-Murcko骨架和化学键构建官能团超图，并引入由电负性和Gasteiger部分电荷驱动的非可逆流。由此产生的环流由厄米磁拉普拉斯算子编码，通过磁切比雪夫编码器处理。通过扰动磁相位生成随机视图，并使用InfoNCE目标函数进行训练。在多个ADMET基准测试中，ChemHyperMag在更少标注样本下表现优于近期方法，且无需构象采样。该模型具备可扩展性，并通过磁相位提供可解释的方向性信号。
原文摘要 · Abstract (English)
Accurate prediction of ADMET (Absorption, Distribution, Metabolism, Excretion, and Toxicity) is important for drug discovery. Most predictors use undirected molecular graphs and pairwise edges. This choice misses asymmetric interactions, nonreversible dynamics, and motif level effects from functional groups and ring systems. We propose ChemHyperMag for multitask ADMET prediction under missing labels. ChemHyperMag builds a functional group hypergraph from rings, BRICS fragments, Bemis-Murcko scaffolds, and bonds. It also defines a potential driven nonreversible flow guided by electronegativity and Gasteiger partial charges. The resulting circulation is encoded by a Hermitian magnetic Laplacian and processed with a magnetic Chebyshev encoder. We perturb magnetic phases to form stochastic views and train with an InfoNCE objective. Experiments on multiple ADMET benchmarks show improvements over recent methods with fewer labeled samples and no conformers. ChemHyperMag is scalable and provides interpretable directional signals through its magnetic phases.
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。