FastMI-HGNet: A Two-Stream Heterogeneous Graph Neural Network for Multi-Omics Disease Classification
Xufeng Fu, Bipeng Lai, Rongling He, Xiuji Huang, Qin Jiang, Hui Li · 2026
Read FastMI-HGNet: A Two-Stream Heterogeneous Graph Neural Network for Multi-Omics Disease Classification by Xufeng Fu, Bipeng Lai, Rongling He, Xiuji Huang, Qin Jiang, Hui 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/PMC13512668.txt.
Description
Xufeng Fu, Bipeng Lai, Rongling He, Xiuji Huang, Qin Jiang, and Hui Li present a two-stream graph neural network for sorting diseases from multi-omics data, in 2026 work. Layers such as genes, RNA, and methylation differ in scale and grain, which makes joining them hard. The design builds a mixed graph of samples and features, with fast mutual information setting the prior strength of dependency edges.
One stream, built on a transformer, reads feature interplay in vector form, while a graph attention stream models structure among samples and molecular features. An uncertainty-aware ensemble steadies answers when samples are few or noisy. Five public benchmarks tested the system, covering Alzheimer data, glioma, breast cancer, colon tumor staging, and a kidney set used as a ceiling check. Sorting held competitive, and explanation tools surfaced breast cancer marker genes. The authors stress readable biomarker ranking as the real prize.
Catalog record
| CREATOR | Xufeng Fu, Bipeng Lai, Rongling He, Xiuji Huang, Qin Jiang, Hui Li |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 77447 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |