Knowledge Graph modulated Deep Learning for Limited-Sample Clinical Data Analysis
Yuwei Xue, Sakib Mostafa, James Zou, Joseph Liao, Maximilian Diehn, Ash A. Alizadeh, Lei Xing, Md Tauhidul Islam · 2026
Read Knowledge Graph modulated Deep Learning for Limited-Sample Clinical Data Analysis by Yuwei Xue, Sakib Mostafa, James Zou, Joseph Liao, Maximilian Diehn, Ash A. Alizadeh, Lei Xing, Md Tauhidul Islam (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/arxiv-2605-24162.txt.
Description
Xue and seven co-authors introduce Graph-in-Graph, letting clinical models use biological knowledge graphs directly, not flat vectors. Each patient becomes a modular graph, with curated pathway graphs setting the edges and gene-expression measurements setting the node features. Across nearly 9700 patients and five tasks, it beats standard rivals, with the widest gains where samples are scarce. On prostate diagnosis it lifts macro-F1 by up to 49 points, and shuffled-graph controls confirm the real pathway wiring drives the gain.
Catalog record
| CREATOR | Yuwei Xue, Sakib Mostafa, James Zou, Joseph Liao, Maximilian Diehn, Ash A. Alizadeh, Lei Xing, Md Tauhidul Islam |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 67075 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.24162