# Trustworthy AI for disease pathogenesis and precision medicine: A disease-agnostic framework from data to deployment

Researcher: Ge Zhang
Chinese name: 张格
ORCID identity anchor: https://orcid.org/0000-0002-3116-3246
Canonical researcher: https://drgezhang.com/#person

Journal: Genes & Diseases
Year: 2026
Type: Article
DOI: 10.1016/j.gendis.2026.102480
Canonical page: https://drgezhang.com/papers/doi-10-1016-j-gendis-2026-102480.html
Markdown record: https://drgezhang.com/papers/doi-10-1016-j-gendis-2026-102480.md

## Full Authors

1. Hao Chi
2. Xin Liu
3. Yalan Yan
4. Lexin Wang
5. Peng Luo
6. Lanqian Su
7. Lai Jiang
8. Xuancheng Zhou
9. Gang Huang
10. Yuxuan Jiang
11. Shiyang Yin
12. Ke Su
13. Haiqing Chen
14. Ge Zhang
15. Pengpeng Zhang
16. Hao Zhang
17. Bicheng Ye
18. Qiang Cao
19. Jie Liu
20. Vedbar Khadka
21. Hua Yang
22. Guanhu Yang
23. Jiayu Xu
24. Peng Wang
25. Xuelian Wang
26. Zhijia Xia
27. Xiaosong Li
28. Youping Deng
29. Zhenjiang Li
30. Ke Xu

## About this record

This is an author-controlled publication record in the Ge Zhang Academic Hub. It identifies this work as part of Ge Zhang’s publication record via ORCID; the publisher version remains the version of record.

<a id="limited-evidence"></a>
## Author evidence note

### Principal contribution

This review connects biomedical data provenance and endpoint design with model selection, causal interpretation, validation, and deployment governance. Its disease-agnostic framework includes an evidence ladder for positioning AI findings and a proposed model-and-data card for reproducible reporting. It synthesizes methods across omics, imaging/pathology, clinical records, and longitudinal data to guide trustworthy research and translation.

Source: [Introduction, manuscript lines 121-133, physical p.9; reporting/model-and-data-card, lines 666-679, physical p.34; evidence ladder, lines 491-518, physical pp.26-27.](https://orca.cardiff.ac.uk/id/eprint/189812/1/1-s2.0-S2352304226004526-main.pdf)

### Evidence ladder

The review organizes support for an AI finding into association, robustness, biological coherence, causal support, perturbation evidence, and clinical utility. The ladder helps researchers identify what additional evidence is needed to move a finding toward a biological explanation or clinical application.

Source: [Evidence ladder, manuscript lines 491–518, physical PDF pages 26–27.](https://orca.cardiff.ac.uk/id/eprint/189812/1/1-s2.0-S2352304226004526-main.pdf)

### Research use

The framework can inform the design and reporting of studies that combine biomedical data types. Its proposed model-and-data card gives researchers a place to document data provenance, model choices and validation conditions alongside the intended use.

Source: [Introduction, manuscript lines 121–133, physical PDF page 9; reporting and model-and-data card, lines 666–679, page 34.](https://orca.cardiff.ac.uk/id/eprint/189812/1/1-s2.0-S2352304226004526-main.pdf)

### Publication and source version

Genes & Diseases made this review available online on 18 September 2026 (article 102480). The Cardiff repository copy is the Journal Pre-proof under CC BY 4.0 and carries the same formal DOI.

Source: [Repository cover and Journal Pre-proof front matter, physical PDF pages 1–3; publisher online date and article number.](https://orca.cardiff.ac.uk/id/eprint/189812/1/1-s2.0-S2352304226004526-main.pdf)

### Evidence scope and limitations

Author-maintained explanation of a published methodological review, based on the publisher's Journal Pre-proof in Cardiff University's repository and official bibliographic records.

- The article synthesizes methods and proposes a framework. Its data-availability statement reports no newly generated or analyzed data; the review does not report original clinical model validation.

### Provenance

- [Cardiff University repository full text](https://orca.cardiff.ac.uk/id/eprint/189812/1/1-s2.0-S2352304226004526-main.pdf) — Publisher Journal Pre-proof; CC BY 4.0. Front matter; manuscript lines 121–133, 491–518, 666–679 and 738–740. Physical PDF pagination includes repository and publisher cover pages.
- [Publisher-deposited citation metadata](https://api.crossref.org/works/10.1016/j.gendis.2026.102480) — Crossref metadata. Formal DOI, journal, 30-author byline, issued year, article number and version-of-record license metadata.

Source review: 2026-11-04.

## Cite this paper

Hao Chi, Xin Liu, Yalan Yan, Lexin Wang, Peng Luo, Lanqian Su, Lai Jiang, Xuancheng Zhou, Gang Huang, Yuxuan Jiang, Shiyang Yin, Ke Su, Haiqing Chen, Ge Zhang, Pengpeng Zhang, Hao Zhang, Bicheng Ye, Qiang Cao, Jie Liu, Vedbar Khadka, Hua Yang, Guanhu Yang, Jiayu Xu, Peng Wang, Xuelian Wang, Zhijia Xia, Xiaosong Li, Youping Deng, Zhenjiang Li, Ke Xu. Trustworthy AI for disease pathogenesis and precision medicine: A disease-agnostic framework from data to deployment. Genes & Diseases. 2026; article 102480. https://doi.org/10.1016/j.gendis.2026.102480

- [BibTeX](https://drgezhang.com/citations/doi-10-1016-j-gendis-2026-102480.bib)
- [RIS / EndNote](https://drgezhang.com/citations/doi-10-1016-j-gendis-2026-102480.ris)
- [CSL JSON](https://drgezhang.com/citations/doi-10-1016-j-gendis-2026-102480.csl.json)

## Links

- DOI: https://doi.org/10.1016/j.gendis.2026.102480
- HTML page: https://drgezhang.com/papers/doi-10-1016-j-gendis-2026-102480.html
- Google Scholar query: https://scholar.google.com/scholar?q=%22Trustworthy%20AI%20for%20disease%20pathogenesis%20and%20precision%20medicine%3A%20A%20disease-agnostic%20framework%20from%20data%20to%20deployment%22
- ORCID: https://orcid.org/0000-0002-3116-3246
