PhenClust

PhenClust clusters biological phenotypes using semantic similarity derived from the UMLS metathesaurus to enable analysis of disease- and drug-associated phenotype patterns.

Key Features:

• Semantic similarity application: Calculates semantic similarity measures derived from the UMLS metathesaurus to group related phenotypes.
• Dockerized deployment: Distributed as Docker containers that include stable installations of two versions of the UMLS metathesaurus.
• Stability and reproducibility: Dockerized deployment supports stable and reproducible computational environments for phenotype clustering.
• Application to drug networks and targets: Summarizes phenotypes linked to specific drugs or drug networks and supports identification of disease clusters and meta-analyses of drug target candidates.
• Facilitation of high-throughput analysis: Processes large sets of phenotype associations from high-throughput experiments to produce clustered summaries.

Scientific Applications:

• Pharmacogenomics: Clusters phenotype data related to drug interactions to aid identification of therapeutic targets and drug-associated disease patterns.
• Systems biology: Groups phenotypes to reveal disease mechanisms and pathway-level relationships.
• Drug discovery and target prioritization: Identifies disease clusters and supports meta-analyses of drug target candidates.
• High-throughput phenotype interpretation: Reduces complexity of large-scale experimental phenotype outputs through clustering.

Methodology:

Computes semantic similarity measures from the UMLS metathesaurus and performs clustering of phenotype data; summarizes phenotypes linked to drugs and drug networks and supports meta-analysis of drug target candidates; provided as Docker containers including two UMLS metathesaurus versions.