Associating expression and genomic data using co-occurrence measures.

Area of Science:

• Genomics
• Bioinformatics
• Cancer Research

Background:

• Exponential increase in omics data necessitates advanced integration strategies.
• Understanding genomic disease mechanisms, like breast cancer, requires multi-omics data fusion.
• Integrating diverse omics data (expression, mutation, copy number) remains a significant challenge.

Purpose of the Study:

• To propose an intuitive workflow for integrative analysis of breast cancer omics data.
• To identify and characterize distinct expression 'regimes' in key breast cancer genes.
• To link gene expression patterns with genomic aberrations for a comprehensive view.

Main Methods:

• Developed an integrative analysis workflow for expression, mutation, and copy number data.
• Analyzed METABRIC and TCGA-BRCA datasets.
• Utilized gene expression 'regime' co-occurrence as an association measure.

Main Results:

• Identified two expression 'regimes' in breast cancer genes, containing significant clinical information.
• Validated gene co-occurrence as a robust association measure across datasets.
• Demonstrated the linkage between expression regimes and genomic aberrations (e.g., MLPH-FOXA1).

Conclusions:

• The proposed workflow facilitates intuitive integration of multi-omics data.
• Gene expression regimes and their co-occurrence offer insights into breast cancer biology.
• Linking expression patterns to genomic changes provides a more complete understanding of breast cancer.