Biomarker Panel Result Class

Stores a fitted biomarker panel including selected biomarkers, performance summaries, and the trained model. This class is returned by fit_panel() and serves as input to evaluate_panel().

Slots

base_features

Character vector of original (untransformed) biomarker names. These are the features selected by the optimization algorithm before any transformation is applied.

features

Character vector of transformed biomarker names used for model fitting. For pairwise transforms, these are the ratio names (e.g., "A–B").

metrics

Named numeric vector summarizing sensitivity, specificity, etc.

objectives

Data frame describing objective values per candidate solution.

control

List of optimization control parameters including feature_transform.

training_data

Signature of the training dataset (e.g., sample and assay info).

model

Fitted model object (e.g., glm or cv.glmnet) trained on the transformed features.