ConsensusClusterPlus

ConsensusClusterPlus implements consensus clustering to provide quantitative and visual stability evidence for estimating the number of intrinsic classes or clusters within datasets, particularly in cancer research.

Key Features:

• Quantitative stability evidence: Computes measures of clustering stability to support selection of the optimal number of clusters.
• Item tracking and item-consensus plots: Provides item tracking and item-consensus plots to follow individual items across different clustering solutions and assess membership consistency.
• Cluster-consensus plots: Generates cluster-consensus plots that visualize agreement between multiple clustering runs.

Scientific Applications:

• Cancer subtype discovery: Estimates intrinsic groups within cancer datasets by identifying stable clusters that may correspond to biologically meaningful subtypes.

Methodology:

Performs multiple runs of a clustering algorithm on resampled datasets to generate a consensus matrix reflecting stability and agreement of cluster assignments across runs, and provides additional visualizations and quantitative metrics such as item tracking, item-consensus, and cluster-consensus plots.