Fig. 4: Experimental validation of ViroTreat predictions. | Communications Biology

Fig. 4: Experimental validation of ViroTreat predictions.

From: A model for network-based identification and pharmacological targeting of aberrant, replication-permissive transcriptional programs induced by viral infection

Fig. 4

a Representative immunofluorescence images of non-infected (Mock) Caco-2 cells, vehicle control (DMSO) treated and SARS-CoV-2 infected cells, and representative examples of a drug showing significant antiviral effect (Cyclosporine), of a drug showing non-significant antiviral effect (Thalidomide) and a drug showing non-significant antiviral effect and cell toxicity (Fedratinib). Drug concentration (μM) is indicated to the left of the images showing triplicated experiments. Cells were stained with DNA dye Draq5 (red) and anti-dsRNA antibody (green). b Scatterplot showing the ViroTreat results (x-axis) compared to the specific antiviral effect (y-axis) experimentally evaluated in Caco-2 colon adenocarcinoma cells. The vertical and horizontal dashed lines represent the thresholds for statistical significance for ViroTreat (p-value = 10−5, BC) and specific antiviral effect (FDR = 0.05), respectively. c ROC analysis for the ViroTreat predictions, considering as positive response a specific antiviral effect at FDR < 0.05 with at least 20% reduction in virus replication. Estimated AUC, 95% confidence interval (CI) and p-value are indicated in the plot. d Effect of 8 drugs, showing the strongest reduction in SARS-CoV-2 replication in Caco-2 cells, on cell viability and SARS-CoV-2 replication in GI organoid-derived 2D primary cell cultures. Bars indicate the mean ± SEM. Antiviral effect: *FDR < 0.05, **FDR < 0.01. Source data in Supplementary Data 4 and 5.

Back to article page