runibic

runibic implements the UniBic algorithm to identify trend-preserving biclusters in gene expression matrices using a longest common subsequence (LCS) framework.

Key Features:

• Trend-Preserving Biclusters Identification: Detects trend-preserving biclusters that represent groups of genes exhibiting consistent expression trends across subsets of samples.
• Longest Common Subsequence (LCS) Framework: Applies an LCS framework to selected pairs of rows within an index matrix derived from the input data matrix to identify seeds for biclusters.
• Performance and Scope: Evaluated on synthetic and real datasets, outperforming several existing biclustering algorithms for broad biclusters while BicSPAM is reported to be superior for narrow biclusters.

Scientific Applications:

• Gene Function Analysis: Identifies groups of co-expressed genes with similar expression trends to infer functional relationships and interactions.
• Disease Research: Reveals gene expression patterns associated with specific conditions or diseases to support studies of mechanisms and biomarkers.
• Data Integration: Handles complex and noisy expression datasets to facilitate integration and comparative analysis across diverse genomic studies.

Methodology:

Constructs an index matrix from the input data matrix and applies the LCS algorithm to selected pairs of rows to identify seeds and extract trend-preserving biclusters.