Mutual information for detecting multi-class biomarkers when integrating multiple bulk or single-cell transcriptomic...

Area of Science:

• Biomedical research
• Bioinformatics
• Genomics
• Transcriptomics
• Systems biology

Background:

• Biomarker detection is crucial in biomedical research.
• Integrating multiple omics studies enhances biomarker discovery power.
• Existing methods are limited to two-class scenarios, not multi-class designs.

Purpose of the Study:

• To develop a statistical framework for detecting biomarkers with concordant multi-class expression patterns across multiple omics studies.
• To address limitations of existing methods in handling multi-class omics data.
• To provide a robust and accurate approach for biomarker discovery in complex biological systems.

Main Methods:

• Proposed Mutual Information Concordance Analysis (MICA) framework.
• Utilized an information-theoretic approach based on mutual information for global testing.
• Implemented post hoc analysis to identify concordant studies for detected biomarkers.
• Applied MICA to transcriptomic and single-cell RNA-Seq data.

Main Results:

• MICA demonstrated improved accuracy and effective false discovery rate control compared to existing multi-class correlation methods in simulations.
• Applied MICA to mouse metabolism and estrogen treatment expression profiles, revealing significant biological insights.
• Identified critical roles of ribosomal function in triple-negative breast cancer microenvironment using single-cell RNA-Seq data.

Conclusions:

• MICA provides a powerful and versatile framework for biomarker detection in multi-class omics studies.
• The method offers enhanced accuracy and robustness for integrating diverse biological datasets.
• MICA has potential for discovering novel therapeutic targets and understanding complex diseases.