To identify potential therapeutic stop-gaps for SARS-CoV-2, we evaluated a library of 1,670 approved and reference compounds in an unbiased, cellular image-based screen for their ability to suppress the broad impacts of the SARS-CoV-2 virus on phenomic profiles of human renal cortical epithelial cells using deep learning. In our assay, remdesivir is the only antiviral tested with strong efficacy, neither chloroquine nor hydroxychloroquine have any beneficial effect in this human cell model, and a small number of compounds not currently being pursued clinically for SARS-CoV-2 have efficacy. We observed weak but beneficial class effects of -blockers, mTOR/PI3K inhibitors and Vitamin D analogues and a mild amplification of the viral phenotype with -agonists.

Development of accurate disease models and discovery of immune-modulating drugs is challenged by the immune system's highly interconnected and context-dependent nature. Here we apply deep-learning-driven analysis of cellular morphology to develop a scalable 'phenomics' platform and demonstrate its ability to identify dose-dependent, high-dimensional relationships among and between immunomodulators, toxins, pathogens, genetic perturbations, and small and large molecules at scale. High-throughput screening on this platform demonstrates rapid identification and triage of hits for TGF-β- and TNF-α-driven phenotypes. We deploy the platform to develop phenotypic models of active SARS-CoV-2 infection and of COVID-19-associated cytokine storm, surfacing compounds with demonstrated clinical benefit and identifying several new candidates for drug repurposing. The presented library of images, deep learning features, and compound screening data from immune profiling and COVID-19 screens serves as a deep resource for immune biology and cellular-model drug discovery with immediate impact on the COVID-19 pandemic.

The continued scaling of genetic perturbation technologies combined with high-dimensional assays (microscopy and RNA-sequencing) has enabled genome-scale reverse-genetics experiments that go beyond single-endpoint measurements of growth or lethality. Datasets emerging from these experiments can be combined to construct "maps of biology", in which perturbation readouts are placed in unified, relatable embedding spaces to capture known biological relationships and discover new ones. Construction of maps involves many technical choices in both experimental and computational protocols, motivating the design of benchmark procedures by which to evaluate map quality in a systematic, unbiased manner. In this work, we propose a framework for the steps involved in map building and demonstrate key classes of benchmarks to assess the quality of a map. We describe univariate benchmarks assessing perturbation quality and multivariate benchmarks assessing recovery of known biological relationships from large-scale public data sources. We demonstrate the application and interpretation of these benchmarks through example maps of scRNA-seq and phenomic imaging data.

The combination of modern genetic perturbation techniques with high content screening has enabled genome-scale cell microscopy experiments that can be leveraged to construct maps of biology. These are built by processing microscopy images to produce readouts in unified and relatable representation space to capture known biological relationships and discover new ones. To further enable the scientific community to develop methods and insights from map-scale data, here we release RxRx3, the first ever public high-content screening dataset combining genome-scale CRISPR knockouts with multiple-concentration screening of small molecules (a set of FDA approved and commercially available bioactive compounds). The dataset contains 6-channel fluorescent microscopy images and associated deep learning embeddings from over 2.2 million wells that span 17,063 CRISPR knockouts and 1,674 compounds at 8 doses each. RxRx3 is one of the largest collections of cellular screening data, and as far as we know, the largest generated consistently via a common experimental protocol within a single laboratory. Our goal in releasing RxRx3 is to demonstrate the benefits of generating consistent data, enable the development of the machine learning methods on this scale of data and to foster research, methods development, and collaboration.

No abstract