multiClust

multiClust integrates multiple gene selection and clustering methodologies in an R-package to evaluate combinations of methods on transcriptomic data for identifying clinically relevant patient subgroups linked to clinical outcomes.

Key Features:

• Integrative Approach: Combines multiple gene selection and clustering methodologies to test different method combinations on transcriptomic datasets.
• Methodological Flexibility: Supports a range of gene selection and clustering approaches so analyses can be tailored to dataset characteristics.
• Variance-Based Ranking Efficacy: Implements variance-based gene ranking and empirically evaluates its performance using clinical outcome data, showing effectiveness when an appropriate number of genes is selected.
• Comprehensive Support for Feature Selection: Provides extensive support for feature selection processes critical for identifying relevant genes prior to clustering.

Scientific Applications:

• Clinical Subgroup Identification: Identifies clinically relevant patient subgroups from transcriptomic profiles by evaluating clusters in relation to clinical outcomes.
• Gene Selection and Clustering Research: Enables comparative studies of gene selection and clustering methodologies to advance understanding of transcriptomic patterns associated with disease.

Methodology:

Integrates and evaluates combinations of gene selection and clustering methods on transcriptomic data with empirical assessment against clinical outcome data, including the use of variance-based gene ranking.