Mining for coexpression across hundreds of datasets using novel rank aggregation and visualization methods.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Gene expression data is crucial for understanding biological processes.
• Analyzing large-scale gene expression datasets presents significant computational challenges.
• Identifying reliable coexpression patterns across multiple experiments is essential for biological insight.

Purpose of the Study:

• To introduce the Multi-Experiment Matrix (MEM), a novel web resource for gene expression similarity searches.
• To provide a tool for integrating and analyzing diverse microarray datasets.
• To facilitate the discovery of robust coexpression patterns across multiple experiments.

Main Methods:

• Utilized rank aggregation to merge information from multiple microarray datasets.
• Developed algorithms for simultaneous statistical significance estimation.
• Implemented automatic detection, characterization, and visualization of coexpression patterns.

Main Results:

• Created a comprehensive web resource (MEM) for cross-dataset gene expression similarity searches.
• Successfully integrated large collections of microarray datasets.
• Enabled identification and visualization of significant coexpression patterns within and across datasets.

Conclusions:

• MEM offers a powerful and accessible platform for exploring gene expression similarities.
• The rank aggregation approach effectively merges information from heterogeneous datasets.
• MEM facilitates the discovery of biologically relevant coexpression networks.