clustComp

clustComp compares and visualizes relationships between clustering results by constructing weighted bi-graphs whose nodes represent clusters and whose edge weights reflect shared-element intersections to assess overlap in genomics and molecular biology.

Key Features:

• Weighted bi-graph representation: Represents clusters as nodes in a weighted bi-graph where edges connect nodes with non-empty intersections and edge weights equal the number of shared elements.
• Overlap visualization: Uses edge thickness and graph layout to visualize the degree of overlap and similarity between clustering outcomes.
• Bioconductor integration: Interoperates with Bioconductor packages and data structures for high-throughput genomic analyses.
• R implementation: Implemented in R to leverage statistical and graphical libraries for cluster comparison and visualization.

Scientific Applications:

• Genomic data analysis: Compares clustering results from genomic datasets to validate consistency of gene expression patterns across conditions or experiments.
• Molecular biology studies: Analyzes overlapping clusters to investigate protein-protein interactions and functional groupings derived from experimental techniques.
• Interdisciplinary research: Provides a framework for comparing clustering outcomes across datasets and methods in cross-disciplinary studies.

Methodology:

Constructs weighted bi-graphs where nodes represent clusters from different clustering algorithms or datasets, draws edges between nodes with shared elements, and assigns edge weights proportional to intersection sizes for visual and quantitative assessment.