arXiv：cs.LG· Yiming Ren, Xiang Liu, Mustafa Hajij, Pietro Li\`o, Guo-Wei Wei·· 4 小时前AI 评分32
数学不变量驱动的拓扑神经网络 MITNN 用于分子与材料性质预测
Mathematical Invariant-Enabled Topological Neural Networks for Molecular and Materials Property Prediction
AI 导读
研究者提出数学不变量驱动的拓扑神经网络（MITNN），将拓扑、谱理论、交换代数、微分几何与离散曲率等多尺度不变量与拓扑神经架构结合，用于分子与材料性质预测。在蛋白质-配体结合、金属有机框架性质、突变诱导的蛋白质溶解度及分子毒性预测任务上，MITNN 均优于现有方法。系统性不变量子集与架构子集分析表明，其性能取决于数学表示与神经架构的配对方式，精选组合优于单一模型及全部组件的简单聚合。
正文
Abstract:Existing molecular and materials learning approaches often rely on a limited set of structural representations, which may capture only selected aspects of complex three-dimensional structure. Here, we introduce mathematical invariant-enabled topological neural networks (MITNNs), a framework that represents complex structures through multiple complementary mathematical views and integrates them with topological neural architectures. MITNNs combine multiscale invariants from topology, spectral theory, commutative algebra, differential geometry, and discrete curvature, capturing complementary structural information from the same system. Systematic invariant-subset, architecture-subset, and ensemble analyses show that predictive performance depends on how mathematical representations and neural architectures are paired, with selected combinations outperforming individual models and the aggregation of all available components. Across protein-ligand binding, metal-organic framework properties, mutation-induced protein solubility, and molecular toxicity prediction, MITNN consistently outperforms existing methods. These results establish MITNN as a mathematically multimodal framework for scientific machine learning.
| Subjects: | Biomolecules (q-bio.BM); Machine Learning (cs.LG) |
| Cite as: | arXiv:2610.07712 [q-bio.BM] |
| (or arXiv:2610.07712v1 [q-bio.BM] for this version) | |
| https://doi.org/10.48550/arXiv.2610.07712 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Yiming Ren [view email]
[v1]
Tue, 6 Oct 2026 04:06:38 UTC (8,621 KB)
来源：arXiv：cs.LG · arxiv.org