Drug discovery is a challenging endeavor – fewer than 10% of drug candidates ever enter the clinic. A major contributor to this is drug discovery teams’ imperfect understanding of the underlying disease pathophysiology. Recently, however, advances in low-cost, high-throughput microscopy methods such as Cell Painting promise to unlock the therapeutic insights currently hidden in cell morphology data. While public datasets for image-based analysis have been difficult to come by historically, the JUMP-Cell Painting Consortium aims to fix that: combining images from nearly 120,000 chemical and 18,000 genetic perturbations and high dimensional data analytics, they’re making available the largest open-source Cell Painting dataset to date and enabling not only retrospective understanding, but also predictive ability for compound activity and toxicity, drug/disease state matching, and more.
Cell Painting uses a set of fluorescent dyes to quantify various biomacromolecules and cell compartments of individual cells, such as DNA, RNA, nucleoli mitochondria, actin, endoplasmic reticulum, Golgi apparatus, and cell membranes in high-throughput plate-based experiments.
Algorithms then process these raw images into hundreds of measurements about each cell, such as the size and shape of each cellular compartment. These data can then be used to directly quantify the morphological changes caused by genetic and chemical perturbations - and subsequently infer for a given disease, which pathways to target and which chemical matter to target them with.
Instead of spending months optimizing a bespoke image-based readout, Cell Painting provides a robust, off-the-shelf solution for phenotype discovery and screening, target + lead compound identification and prioritization, and mechanism of action dissection.
Although JUMP-CELL is generating the raw data and algorithms, confidently interrogating these data requires deep bioinformatics expertise and powerful computing systems. If your team wants to use datasets like JUMP-Cell to better prosecute your therapeutic hypotheses, reach out to Watershed today to speak with our team of expert biologists, bioinformaticians, and data scientists.