Abstract:Cell-based medicinal products (CBMPs) are a growing class of therapeutics that promise new treatments for complex and rare diseases. Given the inherent complexity of the whole human cells comprising CBMPs, there is a need for robust and fast analytical methods for characterization, process monitoring, and quality control (QC) testing during their manufacture. Existing techniques to evaluate and monitor cell quality typically constitute labor-intensive, expensive, and highly specific staining assays. In this work, we combine image-based deep learning with flow imaging microscopy (FIM) to predict cell health metrics using cellular morphology "fingerprints" extracted from images of unstained Jurkat cells (immortalized human T-lymphocyte cells). A supervised (i.e., algorithm trained with human-generated labels for images) fingerprinting algorithm, trained on images of unstained healthy and dead cells, provides a robust stain-free, non-invasive, and non-destructive method for determining cell viability. Results from the stain-free method are in good agreement with traditional stain-based cytometric viability measurements. Additionally, when trained with images of healthy cells, dead cells and cells undergoing chemically induced apoptosis, the supervised fingerprinting algorithm is able to distinguish between the three cell states, and the results are independent of specific treatments or signaling pathways. We then show that an unsupervised variational autoencoder (VAE) algorithm trained on the same images, but without human-generated labels, is able to distinguish between samples of healthy, dead and apoptotic cells along with cellular debris based on learned morphological features and without human input. With this, we demonstrate that VAEs are a powerful exploratory technique that can be used as a process monitoring analytical tool.

Label-free prediction of cell painting from brightfield images

Cell Painting-based bioactivity prediction boosts high-throughput screening hit-rates and compound diversity

Label-free cell cycle analysis for high-throughput imaging flow cytometry

Cell Painting, a high-content image-based assay for morphological profiling using multiplexed fluorescent dyes

Optimizing the Cell Painting assay for image-based profiling

Label-free Detection of Cellular Drug Responses by High-Throughput Bright-Field Imaging and Machine Learning

Live Cell Painting: image-based profiling in live cells using Acridine Orange

Unleashing the potential of cell painting assays for compound activities and hazards prediction

Low concentration cell painting images enable the identification of highly potent compounds

Drug-induced cytotoxicity prediction in muscle cells, an application of the Cell Painting assay

Cell-vision fusion: A Swin transformer-based approach for predicting kinase inhibitor mechanism of action from Cell Painting data

Deep Learning in Label-free Cell Classification

High-Resolution In Silico Painting with Generative Models

Learning representations for image-based profiling of perturbations

Predicting fluorescent labels in label-free microscopy images with pix2pix and adaptive loss in Light My Cells challenge

Phenotypic Profiling of Reference Chemicals across Biologically Diverse Cell Types Using the Cell Painting Assay

Stain-free approach to determine and monitor cell heath using supervised and unsupervised image-based deep learning

DeepPaint: A deep-learning package for Cell Painting Image Classification

In Silico Labeling: Predicting Fluorescent Labels in Unlabeled Images

A Decade in a Systematic Review: The Evolution and Impact of Cell Painting