QUBIC

QUBIC: Qualitative Biclustering of Gene Expression Data

QUBIC performs qualitative biclustering of gene expression matrices to identify subsets of genes exhibiting coordinated expression patterns across specific conditions.

Key Features:

• Computational Optimization: R implementation refactors and optimizes original C code, achieving an average 82% increase in computational efficiency for high-throughput data analysis.
• Discretization: Transforms continuous gene expression data into qualitative discrete representations for biclustering.
• Query-Based Biclustering: Executes targeted biclustering analyses based on user-defined genes or conditions.
• Bicluster Expansion and Comparison: Expands identified biclusters and compares them across datasets or experimental conditions.
• Co-expression Network Analysis: Derives and visualizes gene co-expression networks from biclustered gene sets.

Scientific Applications:

• Transcriptomic Analysis: Identifies condition-specific co-expressed gene subsets to investigate biological processes, regulatory mechanisms, disease pathways, developmental stages, and environmental responses.

Methodology:

QUBIC discretizes continuous gene expression data into qualitative states and applies a biclustering algorithm to detect submatrices enriched for consistent expression patterns across subsets of genes and conditions. Post-processing functions enable bicluster refinement, cross-dataset comparison, and construction of co-expression networks from bicluster-derived gene sets.