High content phenotypic screening is a powerful approach for identifying new genes and compounds that modulate cellular disease phenotypes. Array-based fluorescence imaging platforms are commonly used for phenotypic screening due to their high throughput capabilities. However, the approach has major limitations including well-to-well variability, resource-intensive experimental setup and image analysis, and preference for adherent fixed cells. Here, we introduce a new approach for pooled phenotypic screening by combining morphological profiling with DNA barcoding technology. We show application of the VisionSort platform, a new AI-driven cell sorter powered by Ghost Cytometry, combined with DNA barcoding for label-free high-throughput pooled phenotypic compound screening. VisionSort provides advantages over traditional screening approaches including i) scalability to large screening libraries, ii) applicability to a wide new range of morphological phenotypes, and iii) compatibility with commercial sequencing platforms for DNA barcode readouts.
