LogicalClusteringSu​ite

Logical clustering suite with graphical user interface.

The Logical Clustering Suite (LCS) is a MATLAB-based application for the conceptual clustering of gene expression data. Instead of grouping genes by mutual similarity (as in hierarchical clustering or k-means), LCS defines clusters by experimentally derived logical phenotypes (IPs). Each IP is a binary vector (0/1) representing the expected activation pattern across groups. Genes are assigned to the IP they best match, and their assignments are supported by multiple statistical tests and an integrated ranking system.

Advantages:

- Interpretability: clusters are defined by logical patterns of 1/0 responses, which directly reflect experimental concepts.

- Statistical rigor: each gene/IP assignment is evaluated with correlation distance, negative-binomial testing (BH-FDR), one-way ANOVA, fold change, and a composite Z-score.

- Flexible focusing: Boolean filters and IP keep-lists allow users to restrict analyses to IPs of interest.

- Comprehensive outputs: tables, heatmaps, boxplots, enrichment analysis (sGEA), and network representations.

- Required sheet: SourceData.

- Columns: GeneID, GeneSymbol, optional IDs, then numeric sample columns.

- Sample naming: Group_Replicate (e.g., wt_CTRL_1).