ArrayCluster

ArrayCluster identifies molecular subtypes from DNA microarray gene expression data by grouping samples according to their expression patterns to discover disease-associated subtypes.

Key Features:

• Model-Based Clustering Approach: Employs a model-based clustering method called mixed factors analysis to address clustering in gene expression data.
• Handling High-Dimensional Data: Manages datasets in which the number of genes far exceeds the number of samples to mitigate over-learning and overfitting common in DNA microarray experiments.
• Analytic Tools for Clustering: Provides analytic tools for clustering gene expression data derived from DNA microarray experiments.
• Data Visualization: Offers visualization capabilities to explore and interpret complex gene expression patterns and relationships between molecular subtypes.
• Automatic Detection of Module Transcriptional Relevance: Includes an automatic detector that identifies modules of transcriptionally relevant genes associated with calibrated molecular subtypes.

Scientific Applications:

• Molecular subtype discovery: Identification of previously unknown molecular subtypes of diseases from gene expression profiles.
• Cancer genomics: Subtype identification in cancer studies using DNA microarray expression data.
• Large-scale gene expression studies: Analysis of DNA microarray experiments where dimensionality (genes) greatly exceeds sample size.
• Systems biology and pathway analysis: Detection of transcriptionally relevant gene modules to support systems-level and pathway analyses.
• Pharmacogenomics and therapeutic stratification: Definition of molecular subtypes to inform stratification in pharmacogenomics studies.

Methodology:

Model-based clustering via mixed factors analysis; automatic detection of transcriptionally relevant gene modules associated with calibrated molecular subtypes; clustering of DNA microarray gene expression data.