oCEM

oCEM identifies and analyzes overlapping co-expressed gene modules in gene expression datasets using decomposition methods to capture module overlap and local co-expression within subsets of biological samples, implemented as an R package.

Key Features:

• Overlapping module detection: Automatic detection and analysis of overlapping co-expressed gene modules.
• Local co-expression modeling: Addresses local co-expression patterns within subsets of biological samples.
• Decomposition methods: Leverages decomposition methods for module identification and representation.
• Unsupervised clustering: Uses unsupervised clustering approaches tailored for both sample- and gene-clustering tasks with a focus on gene-clustering.
• Permutation-based component selection: Implements an auxiliary statistical permutation procedure to determine the optimal number of principal components.

Scientific Applications:

• Identification of biologically relevant modules: Identifies co-expressed gene modules that are biologically relevant from gene expression data.
• Analysis of local co-expression: Reveals and characterizes local co-expression patterns within subsets of biological samples.

Methodology:

Employs decomposition methods and unsupervised clustering for sample- and gene-clustering (with emphasis on gene-clustering) and uses a permutation-based statistical procedure to select the optimal number of principal components.