BreastSubtypeR

BreastSubtypeR unifies multiple published intrinsic molecular subtyping classifiers into a single, reproducible framework for breast cancer research, enabling systematic cross-method comparison and robust cohort-aware selection.

Key Features:

• AUTO mode: Quantitatively evaluates cohort composition and selectively activates classifiers whose assumptions are met, reducing user bias and improving robustness across skewed or subtype-specific cohorts.
• Multi-method harmonization: Integrates several published nearest-centroid (NC) and single-sample predictor (SSP) classifiers with method-specific preprocessing and robust probe-to-gene mapping.
• Unified normalization pipeline: Provides a standardized, method-specific normalization workflow within a Bioconductor framework.

Scientific Applications:

• Intrinsic molecular subtyping: Enables standardized classification of breast cancer samples into molecular subtypes for understanding disease biology and guiding personalized treatment.
• Cross-method comparison: Facilitates systematic comparison of multiple intrinsic subtyping classifiers within a single analysis framework.
• Cohort-aware classifier selection: Supports robust subtype prediction across skewed or subtype-specific cohorts by matching classifier assumptions to cohort composition.

Methodology:

BreastSubtypeR implements nearest-centroid (NC) and single-sample predictor (SSP) classification algorithms with method-specific preprocessing, optimized probe-to-gene mapping, and cohort-aware AUTO mode evaluation.