3Mont: A multi-omics integrative tool for breast cancer subtype stratification

Area of Science:

• Genomics and Bioinformatics
• Cancer Research
• Computational Biology

Background:

• Breast cancer (BRCA) is a heterogeneous disease with diverse molecular sub-types, necessitating tailored treatment strategies.
• Hormone Receptor-negative (HR-) BRCA, particularly Basal-like BRCA, exhibits aggressive tumor growth and poorer prognosis compared to HR+ subtypes.
• Accurate sub-typing and biomarker identification are crucial for improving survival rates and predicting prognosis in BRCA patients.

Purpose of the Study:

• To introduce a novel computational tool, 3-Multi-Omics Network and Integration Tool (3Mont), for integrating multi-omics data in BRCA research.
• To develop a machine learning-based approach for identifying prominent biomarkers and stratifying BRCA sub-types.
• To enhance the efficiency and analytical capabilities for understanding complex biomarker interactions in cancer.

Main Methods:

• The 3Mont tool integrates multi-omics data using a grouping function to detect 'pro-groups' and assigns scores via Feature Importance Scoring (FIS).
• Machine learning models are built upon these pro-groups to extract biomarkers for distinguishing BRCA sub-types.
• The FIS component equalizes feature numbers across pro-groups, achieving a 20% speedup compared to the 3Mint tool.

Main Results:

• 3Mont successfully integrates diverse -omics data, enabling the identification of key biomarkers for BRCA sub-type stratification.
• The tool generates networks illustrating the interplay of prominent biomarkers across different omics layers.
• The approach facilitates a deeper understanding of biomarker dynamics and their concerted actions within biological groups.

Conclusions:

• 3Mont provides a novel and efficient method for analyzing multi-omics data to stratify breast cancer sub-types.
• The generated biomarker networks offer insights into the complex molecular mechanisms driving different BRCA subtypes.
• This tool has the potential to advance the development of personalized treatment strategies for breast cancer patients.