CLEAN: CLustering Enrichment ANalysis.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Integrating functional genomics data with biological knowledge is crucial for analyzing large-scale gene expression datasets.
• Cluster analysis is widely used to identify biologically meaningful gene groups and pathways.
• Assessing the functional coherence of gene clusters is key for robust biological interpretation.

Purpose of the Study:

• To develop a computational framework for integrating knowledge-based functional categories with genomics data cluster analysis.
• To introduce a gene-specific functional coherence score (CLEAN score) for improved analysis.
• To enhance the reproducibility and interpretability of cluster analysis results.

Main Methods:

• Developed a computational framework integrating functional categories with cluster analysis.
• Introduced the gene-specific functional coherence score (CLEAN score) based on correlating clustering structure with functional categories.
• Implemented the framework as an R package with routines for calculating scores and visualization tools.

Main Results:

• The CLEAN score improves the reproducibility of conclusions drawn from cluster analysis.
• It differentiates functional coherence levels within clusters based on enriched functional categories.
• This approach yields higher reproducibility across datasets and identifies more informative genes compared to traditional methods.

Conclusions:

• Gene-specific functional coherence scores enhance the reproducibility of conclusions about co-expressed gene clusters.
• This method simplifies comparisons between different clustering algorithms.
• Provides a tool for selecting genes with functionally coherent expression profiles.