JUMP-CP consortium

The JUMP-CP (Joint Undertaking for Morphological Profiling-Cell Painting) Consortium is a global, multi-institutional collaboration initiated by the Broad Institute with the goal of building the world’s largest open-access dataset of cellular images generated through the Cell Painting assay. Its mission is to enable image-based drug discovery by providing standardized, high-quality phenotypic data that can be used to identify compound bioactivity, modes of action, and potential off-target effects across a wide range of chemical and genetic perturbations. https://jump-cellpainting.broadinstitute.org/

The JUMP-CP Data Explorer is a free interactive online platform developed to navigate, search, and visualize the massive image-based dataset produced by the consortium. It allows users to explore Cell Painting data by compound, gene perturbation, phenotypic cluster, or similarity to reference profiles. Researchers can query the dataset for morphological signatures, download image-derived features, and compare phenotypic effects across conditions and time points.

Cell Painting data consists of high-content images of cells stained with a combination of fluorescent dyes that label various organelles and cellular components. The analysis of these images extract thousands of single cell features generating unique morphological profiles that capture the phenotypic response of cells to chemical or genetic perturbations. In drug discovery, these profiles allow researchers to screen compounds in a target-agnostic manner, uncover novel modes of action, and prioritize hits based on functional similarity rather than chemical structure alone.

The data generated by the JUMP-CP Consortium is publicly available and can be accessed via the Broad Institute’s JUMP hub providing access to GitHub repositories, and associated cloud-hosted data portals such as Ardigen phenAID JUMP-CP Data Explorer.  Users can download raw images, extracted features, and metadata, or analyze data directly through cloud-integrated tools provided by the consortium.

Ardigen is a contributing partner to the JUMP-CP Consortium, supporting the initiative through its expertise in AI, bioinformatics, and high-content image analysis. Ardigen develops tools for morphological profiling, phenotype clustering, and prediction of compound mode  of action, and contributes to the interpretation and downstream utility of the public dataset in real-world drug discovery applications.

Image-based screening refers to the use of high-throughput microscopy to analyze cellular responses to genetic or chemical perturbations. Depending on the assay design, it can be used to observe morphological changes (as in Cell Painting) or to monitor the localization or abundance of specific proteins of interest using targeted stains or reporters.

Applications include hit identification, mode-of-action studies, compound clustering, toxicity assessment, and drug repurposing. While image-based screening does not necessarily require prior knowledge of molecular targets, it can also be applied in target-specific contexts.
It is one of several approaches used in phenotypic screening and is often combined with other methods as part of an integrated drug discovery strategy.

The JUMP-CP Consortium includes leading academic institutions, pharmaceutical companies, biotech firms, and nonprofit research organizations. Founding members include the Broad Institute, Bayer, Roche, Merck, Pfizer, AstraZeneca, the Chan Zuckerberg Initiative, and others. Ardigen is part of this global collaboration, contributing to the scientific and technical development of the platform.

The public dataset includes raw cell images, morphological feature sets extracted using CellProfiler and other tools, compound metadata, perturbation annotations, and experimental context such as cell line, time point, and staining protocol. This multimodal data can be used for machine learning, phenotypic similarity analysis, and integrative biology research. 

These AI solutions are overseen by experts like Dawid Rymarczyk, PhD. He is Director of AI Solutions at Ardigen. His expertise in deep learning and image analysis is crucial for developing these cutting-edge platforms.