Taking Cell Phenotype Analysis to New Levels of Detail and Automation

Nuclei counting is one of the most common cell phenotype analysis tasks for biological research. Automating it is crucial for many applications and further downstream analyses. This workflow is based on a pre-trained Deep Learning model, automatically segmenting, separating, and counting nuclei based on any fluorescent nuclear marker, such as DAPI, in one or more microscopy images. It supports time-series images as well as multi-well measurements. The output is a matrix that shows the number of nuclei per image in each well per time point (if applicable). The respective Deep Learning model is trained on datasets from multiple microscopes with different resolutions and magnifications.

Phenotypic screening is a target agnostic approach to drug discovery that monitors for phenotypic changes in cells. This application allows for high-throughput quantitative analysis of multi-well plates providing outputs on various sub-cellular intensities and morphological measurements on complex cellular models. In this example, a nuclear maker is used to identify all cells and specific cell populations are identified based on nuclear and cytoplasmic marker combinations. Cytoplasmic markers are also used to characterize the shape and size of all individual cells expressing this marker.