machine-learning-for-omics

To perform biomarker discovery from omics data, you need a feature matrix and labels. The workflow trains a predictive model, outputs validation metrics, and provides interpretable explanation summaries for the identified biomarkers.

Yes, survival analysis and classification are supported within the same omics workflow. The Skill handles supervised learning tasks for both disease status classification and patient survival prediction using your defined feature matrix and outcomes.

Supervised learning on omics-derived features requires an omics feature matrix, corresponding labels or patient outcomes, and a defined data split design. These inputs yield a trained model, validation metrics, and explanation summaries.

Model explainability for omics predictive modeling works by generating interpretation summaries alongside trained models and validation metrics. This ensures that the biomarker discovery and risk prediction processes remain interpretable and trustworthy.

Yes, the workflow supports robust evaluation and calibration checks for omics data. You can define data splits and perform calibration checks to ensure trustworthy results for your trained predictive models and biomarker discovery outputs.