Representative applications in cancer, prognosis, and drug response.
| Application domain | Representative studies | Methods | Omics layers | XAI contribution | Clinical relevance | References |
|---|---|---|---|---|---|---|
| Cancer subtype classification | AMMO; DeepMoIC; MoAGNN | Attention-fusion; GCNs; hierarchical GNN | Genomics, transcriptomics, clinical | Attention weights identify key genes and pathways driving subtypes | Guides treatment selection; identifies targetable subtypes | [63, 86, 100] |
| Biomarker discovery | DeepKEGG; SHAP-based analyses | Pathway-informed DL; explainable GNNs | Multi-omics (varies) | Feature attribution prioritizes genes, pathways for validation | Novel therapeutic targets; diagnostic signatures | [53, 71, 72] |
| Drug response prediction | DeepFusionCDR; DeepInsight-3D | Transformers; CNNs; DL | Genomics, transcriptomics, proteomics | SHAP, attention identify resistance/sensitivity mechanisms | Treatment selection; avoidance of ineffective therapies | [90, 91, 101] |
| Survival/Prognosis | Autosurv; CoFormerSurv; Pathformer | Interpretable DL; collaborative transformers; pathway-informed | Multi-omics + clinical | Time-dependent feature importance; pathway-level explanations | Risk stratification; treatment intensity decisions | [55, 56, 84] |
| Immunotherapy response | Immunotherapy survival XAI; Blood-cancer multi-omics | Explainable ML; SHAP | Multi-omics, immune signatures | Identifies immune-related gene signatures and TMB contributions | Patient selection for checkpoint inhibitors | [70, 89] |
| Single-cell integration | FactVAE | Factorized VAEs; multimodal DL | Single-cell multi-omics | Disentangled latent factors; cell-type specific explanations | Cellular heterogeneity; developmental trajectories | [34, 102] |
XAI: explainable artificial intelligence; GCNs: graph convolutional networks; GNNs: graph neural networks; SHAP: SHapley Additive exPlanations; DL: deep learning; ML: machine learning; VAEs: variational autoencoders.
During the preparation of this work, the authors used Google AI tools to improve language and readability. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article.
MH: Conceptualization, Writing—original draft, Visualization. SH: Writing—review & editing, Supervision. KA: Methodology, Writing—review & editing, Formal analysis. MW: Methodology, Formal analysis, Writing—review & editing. All authors have read and approved the final version of the manuscript.
The authors declare that they have no conflicts of interest.
Not applicable.
Not applicable.
Not applicable.
No new data were generated or analyzed in this review. All referenced datasets are publicly available from the sources cited.
Not applicable.
© The Author(s) 2026.
Open Exploration maintains a neutral stance on jurisdictional claims in published institutional affiliations and maps. All opinions expressed in this article are the personal views of the author(s) and do not represent the stance of the editorial team or the publisher.