CPSM: An R Package for Cancer Patient Survival Risk Model Using Transcriptomics and Clinical Data - PubMed
2 hours ago
- #R-package
- #bioinformatics
- #cancer-survival-modeling
- CPSM is an R package for personalized cancer survival risk modeling using transcriptomics and clinical data, addressing heterogeneity overlooked by traditional methods.
- It offers a reproducible pipeline with 10 core functions across four steps: data preprocessing, feature selection, survival risk-group modeling, and visualization including nomograms.
- The package was validated on TCGA datasets for cancers like glioblastoma, leukemia, pancreatic adenocarcinoma, and breast cancer, handling high-dimensional data with over 60,000 RNA transcripts.
- Evaluation uses repeated cross-validation with uncertainty quantification, ensuring robustness in high-dimensional, small-sample settings.
- CPSM enhances patient-clinician communication through individualized survival curves and is freely available on Bioconductor and GitHub for broader use.