MacroTox: A Macroscopic Graph Topology-Based Multimodal Learning Framework for Robust Molecular Toxicity Prediction
He Huang, Qinyi Wang, Manzhan Zhang, Pan Dou, Xiaobo Yang, Honglin Li, Shiliang Li · 2026
Read MacroTox: A Macroscopic Graph Topology-Based Multimodal Learning Framework for Robust Molecular Toxicity Prediction by He Huang, Qinyi Wang, Manzhan Zhang, Pan Dou, Xiaobo Yang, Honglin Li, Shiliang Li (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13625743.txt.
Description
He Huang and co-authors present MacroTox in a 2026 paper on predicting drug harm to bone. Most toxicity models learn each molecule alone and miss how molecules resemble each other across the whole set. That gap breaks the learned codes and hides toxic cases, which drives missed alarms.
MacroTox builds a live graph of like drugs inside each training batch and fuses it with each molecule's own traits. On a curated bone toxicity set it scored high on both catches and clear calls. Part-removal tests showed the graph step drives the gain, even when data are scarce.
Catalog record
| CREATOR | He Huang, Qinyi Wang, Manzhan Zhang, Pan Dou, Xiaobo Yang, Honglin Li, Shiliang Li |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 50 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |