3Mint

3Mint integrates mRNA, microRNA (miRNA), and DNA methylation data using machine learning to identify molecular subtypes and cross-omics biomarkers in Breast Cancer (BRCA).

Key Features:

• Multi-Omics Integration: 3Mint combines gene expression (mRNA), microRNA (miRNA), and methylation datasets to provide a holistic view of molecular interactions in BRCA.
• Machine Learning-Based Approach: The method employs advanced machine learning algorithms to uncover systematic connections across omics layers and enhance predictive power.
• Grouping and Scoring Mechanism: 3Mint groups and scores biological entities based on integrated multi-omics data while leveraging existing biological knowledge for feature selection.
• Enhanced Diagnostic and Therapeutic Insights: Integrative analyses reveal regulatory mechanisms and candidate targets that inform diagnosis and therapeutic hypotheses for BRCA.
• Performance Evaluation: Computational assessments reported 95% classification accuracy for BRCA molecular subtypes using fewer genes compared to miRcorrNet, and incorporation of methylation data further refines the analysis.

Scientific Applications:

• Cancer Research: Elucidating interactions among genomic, epigenetic, and transcriptomic factors that contribute to tumor development and progression in BRCA.
• Biomarker Discovery: Identifying cross-omics biomarker groups for potential diagnostic and prognostic markers in BRCA.
• Personalized Medicine: Supporting molecular profiling of individual tumors to inform personalized treatment strategies.

Methodology:

An integrative machine learning framework systematically combines and analyzes mRNA, miRNA, and methylation data, groups and scores biological entities using existing biological knowledge, and identifies patterns and interactions to classify BRCA molecular subtypes.