MoReCluster

MoReCluster is an open-source Python framework for grouping and clustering chemicals according to shared molecular response patterns and phenotypic effects. It uses toxicogenomics associations curated in the Comparative Toxicogenomics Database (CTD) to identify substances that exhibit related biological activity.

The framework builds on the Chemical–Gene–Phenotype–Disease (CGPD) tetramer concept, which connects chemical exposures with affected genes, observed phenotypes, and associated diseases. By integrating these relationships, MoReCluster supports the discovery of coherent chemical groups based on biological response evidence rather than chemical structure alone. This can help researchers generate hypotheses about shared molecular mechanisms and related phenotypic outcomes.

Main features

• Biology-driven chemical grouping: Groups chemicals by shared molecular response patterns and phenotypic effects derived from toxicogenomics data.
• CGPD-based representation: Integrates chemical, gene, phenotype, and disease associations into a unified analytical framework.
• Configurable data filtering: Enables users to adjust the evidence included in an analysis according to the scientific question.
• Flexible clustering: Provides configurable strategies for exploring similarities between chemicals.
• Context-aware analyses: Chemical relationships can be investigated across different tissues, species, and data dimensions.
• Reproducible distribution: The published release includes both the source code and a SIF container image suitable for Apptainer/Singularity-compatible environments.

MoReCluster is intended for researchers working in toxicogenomics, computational toxicology, environmental health, and related fields. It provides a reproducible basis for exploring complex relationships between chemical exposures and biological effects and for identifying groups of substances with potentially related modes of action.

MoReCluster is written in Python and distributed under the GNU General Public License v3.0 or later. The software is under active development.

Resources

• Source-code repository
• Software release on Zenodo

Citations

Software package

Canzler, S., Lehmann, J., and Hackermüller, J. (2025). MoReCluster: Clustering chemicals by molecular response patterns and phenotypic effects [Computer software]. Zenodo. https://doi.org/10.5281/zenodo.15188632

Associated publication

Canzler, S., Lehmann, J., Schor, J., Busch, W., Iacono, G., and Hackermüller, J. (2026). From toxicogenomics data to cumulative assessment groups: a framework for chemical grouping. Archives of Toxicology, 100, 173–191. https://doi.org/10.1007/s00204-025-04133-w