Luminal Breast Cancer Subtype Biomarkers from Multi-Omics Machine Learning Analysis

A dataset of 30 methylation-driven biomarkers identified to differentiate luminal A and B breast cancer subtypes. The data was generated by applying the INTEND machine learning algorithm, trained on 4,441 paired RNA-seq and DNA methylation samples from 14 cancer types, to 537 TCGA-BRCA luminal samples. The model identified 2,670 high-confidence genes with strong methylation-expression coupling, achieving an average R² of 0.76.

Use Cases

• Train subtype classification models based on the 30 identified methylation-driven biomarkers.
• Investigate methylation-expression regulatory mechanisms in breast cancer using the filtered gene set.
• Validate findings on epigenetic drivers of luminal breast cancer differentiation.
• Benchmark new multi-omics integration methods against the described INTEND pipeline.

Strengths

• Includes 30 specific biomarkers identified for subtype differentiation.
• Model was trained on a large, independent cohort of 4,441 multi-omics samples to prevent data leakage.
• Analysis achieved a high average predictive accuracy (R² = 0.76).

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for the final biomarker dataset is unknown, which may limit suitability assessment.
• Data may reflect bias inherent to the TCGA-BRCA source cohort.

Provenance

Source

TCGA-BRCA (The Cancer Genome Atlas Breast Invasive Carcinoma) and 14 other cancer types.

Collection Method

Machine learning pipeline (INTEND algorithm) applied to paired RNA-seq and DNA methylation data.

Time Range

null

Freshness

Last updated 2026-05-16 13:30:56; freshness should be verified.

Geography

null