MOG

MOG performs interactive exploratory analysis and visualization of large-scale omics datasets, integrating metadata and statistical methods to detect co-expression, differential expression, and differential correlation patterns.

Key Features:

• Implementation and data handling: Java-based application employing advanced indexing and buffering schemes to manage and analyze large datasets on local machines.
• Multithreading: Multithreading support for efficient processing of large datasets.
• Metadata and ontology integration: Incorporates sample and feature metadata and ontology annotations into analyses.
• Visualizations: Interactive visualizations including line charts, box plots, scatter plots, histograms, and volcano plots.
• Subsetting and grouping: Enables focusing on specific groups of samples or genes based on expression values, statistical associations, metadata terms, and ontology annotations.
• Statistical analyses: Implements co-expression analysis, differential expression analysis, and differential correlation analysis with significance testing.
• R integration: Exports subsets of data to R for further customized analyses.
• Supported data types: Accepts numerical datasets and existing MOG projects and supports RNA-Seq, microarray, and metabolomics data.

Scientific Applications:

• Cancer biomarker discovery: Applied to large curated human cancer RNA-Seq datasets to identify putative biomarker genes across tumor types.
• Plant omics analysis: Applied to Arabidopsis thaliana microarray and metabolomics datasets for exploratory analysis and pattern detection.

Methodology:

Computational methods explicitly include advanced indexing and buffering schemes for efficient data handling, multithreading, co-expression analysis, differential expression analysis, differential correlation analysis with significance testing, and exporting subsets to R.