Transferable Biomarker Panel Result Class

Extends BiomarkerPanelResult with additional slots for train/validation/held-out split strategy and Neyman-Pearson threshold selection for cross-cohort generalizability.

Slots

np_threshold

Numeric threshold selected via Neyman-Pearson classification.

np_alpha

Type I error rate for NP threshold selection.

np_delta

Tolerance parameter for NP threshold selection.

per_cohort_metrics

Data frame with per-cohort performance metrics.

weighted_variance

List containing sensitivity and specificity variance (inverse sample-size weighted across cohorts).

partition_info

List containing partition sizes and ratios.