Deep Transfer Learning Approach for Automatic Recognition of Drug Toxicity and Inhibition of SARS-CoV-2
Julia Werner,
Raphael M. Kronberg,
Pawel Stachura,
Philipp N. Ostermann,
Lisa Müller,
Heiner Schaal,
Sanil Bhatia,
Jakob N. Kather,
Arndt Borkhardt,
Aleksandra A. Pandyra,
Karl S. Lang,
Philipp A. Lang
Affiliations
Julia Werner
Department of Molecular Medicine II, Medical Faculty, Heinrich-Heine-University, 40225 Düsseldorf, Germany
Raphael M. Kronberg
Department of Molecular Medicine II, Medical Faculty, Heinrich-Heine-University, 40225 Düsseldorf, Germany
Pawel Stachura
Department of Molecular Medicine II, Medical Faculty, Heinrich-Heine-University, 40225 Düsseldorf, Germany
Philipp N. Ostermann
Institute of Virology, Medical Faculty, Heinrich-Heine-University, 40225 Düsseldorf, Germany
Lisa Müller
Institute of Virology, Medical Faculty, Heinrich-Heine-University, 40225 Düsseldorf, Germany
Heiner Schaal
Institute of Virology, Medical Faculty, Heinrich-Heine-University, 40225 Düsseldorf, Germany
Sanil Bhatia
Department of Pediatric Oncology, Hematology and Clinical Immunology, Medical Faculty, Center of Child and Adolescent Health, Heinrich-Heine-University, 40225 Düsseldorf, Germany
Jakob N. Kather
Department of Medicine III, University Hospital RWTH Aachen, 52074 Aachen, Germany
Arndt Borkhardt
Department of Pediatric Oncology, Hematology and Clinical Immunology, Medical Faculty, Center of Child and Adolescent Health, Heinrich-Heine-University, 40225 Düsseldorf, Germany
Aleksandra A. Pandyra
Department of Pediatric Oncology, Hematology and Clinical Immunology, Medical Faculty, Center of Child and Adolescent Health, Heinrich-Heine-University, 40225 Düsseldorf, Germany
Karl S. Lang
Institute of Immunology, Medical Faculty, University of Duisburg-Essen, 45147 Essen, Germany
Philipp A. Lang
Department of Molecular Medicine II, Medical Faculty, Heinrich-Heine-University, 40225 Düsseldorf, Germany
Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) causes COVID-19 and is responsible for the ongoing pandemic. Screening of potential antiviral drugs against SARS-CoV-2 depend on in vitro experiments, which are based on the quantification of the virus titer. Here, we used virus-induced cytopathic effects (CPE) in brightfield microscopy of SARS-CoV-2-infected monolayers to quantify the virus titer. Images were classified using deep transfer learning (DTL) that fine-tune the last layers of a pre-trained Resnet18 (ImageNet). To exclude toxic concentrations of potential drugs, the network was expanded to include a toxic score (TOX) that detected cell death (CPETOXnet). With this analytic tool, the inhibitory effects of chloroquine, hydroxychloroquine, remdesivir, and emetine were validated. Taken together we developed a simple method and provided open access implementation to quantify SARS-CoV-2 titers and drug toxicity in experimental settings, which may be adaptable to assays with other viruses. The quantification of virus titers from brightfield images could accelerate the experimental approach for antiviral testing.
