ArrayMining.net

ArrayMining.net performs automated integrative analysis of DNA microarray data to enable cross-study normalization, feature selection, clustering, prediction, co-expression network analysis, and gene set analysis.

Key Features:

• Cross-study normalization: Implements normalization methods to adjust microarray data from multiple studies for comparability across experiments.
• Cross-platform data harmonization: Supports integration and harmonization of datasets derived from different microarray platforms.
• Ensemble learning and consensus clustering: Applies ensemble learning approaches and consensus clustering to aggregate results from multiple algorithms for robust pattern discovery.
• Ensemble feature selection and prediction: Aggregates outcomes from multiple feature selection and prediction algorithms to enhance robustness of biomarker identification and classifiers.
• Co-expression network analysis: Performs analysis of gene co-expression relationships to identify correlated gene modules.
• Gene set analysis: Provides gene set analysis to assess functional annotations and pathway-level associations.
• Automatic parameter selection: Incorporates automatic parameter selection mechanisms to reduce manual tuning of algorithm parameters.
• Modular architecture: Organizes analytical methods into modular components for combining multiple algorithms and workflows.
• Integration with external resources: Links analyses to external web tools and databases for functional annotation and literature mining.

Scientific Applications:

• Meta-analysis and cross-study integration: Enables integration of microarray datasets for meta-analyses and cross-study comparisons.
• Disease-related pattern discovery: Facilitates identification of consistent gene expression patterns and genetic components associated with diseases.
• Robust biomarker discovery and classification: Supports discovery of biomarkers and development of predictive models using ensemble feature selection and prediction.
• Pathway and functional interpretation: Enables exploration of pathways and functional annotations through gene set analysis and linked annotation resources.
• Gene network characterization: Supports characterization of gene co-expression networks and module identification.

Methodology:

Uses cross-study normalization, cross-platform harmonization, ensemble learning, consensus clustering, ensemble feature selection and prediction, co-expression network analysis, gene set analysis, and automatic parameter selection.