A comprehensive approach to identifying patient subgroups with distinct survival rates by leveraging multiomics datasets, namely, microRNA, gene expression, and DNA methylation. This project showcases:
- Integration of heterogeneous big data into intrinsic structures.
- Dimensionality reduction via PCA.
- Graph construction for patient data.
- Data embedding using the Grassmann manifold for patient clustering.
- Extensive testing on datasets from The Cancer Genome Atlas (TCGA).
By integrating these varied datasets on a Grassmann manifold, our method outperforms conventional methods in clustering accuracy and survival rate prediction, paving the way for precise and personalized medical treatments.
Multi-omics, Cancer Subtype, Graphs, Grassmann Manifold, Patient Subgrouping, PCA, Data Integration, The Cancer Genome Atlas, Survival Rates, Clustering, Dimensionality Reduction.
If you find this work beneficial and utilize it in your research, please cite our original article: Patient subgrouping with distinct survival rates via integration of multiomics data on a Grassmann manifold.