BioMorph is a Cell Painting - Cell Health derived feature space designed to enhance the interpretability of cell morphological features. With a commitment to advancing the field of cell morphology analysis, this initiative curates a comprehensive re-organization of data that incorporates biological meaning to CellProfiler features.

Built on the Cell Painting dataset, the BioMorph dataset encompasses many biologically relevant profiles or features for 30,000+ compounds. The resource is publicly available to researchers to assess its performance as a feature space for machine learning models.

For referencing, please cite:
Seal, S, Carreras-Puigvert, J, Carpenter, AE, Spjuth, O, and Bender, A (2023). From Pixels to Phenotypes: Integrating Image-Based Profiling with Cell Health Data Improves Interpretability. bioRxiv, 2023.07.14.549031.


compound profiles from the Cell Painting dataset covering small molecules and FDA-approved drugs


CRISPR perturbations
in Cell Health assays for three cell lines (A549, ES2, & HCC44) used for biological organisation


BioMorph terms
integrate a unique combination of information from perturbations and Cell Health assays

Schema Image

Schema of
BioMorph terms