Fig. 2: Performance of COVID-19 diagnostic classifiers based on patient gene expression. | Nature Communications

Fig. 2: Performance of COVID-19 diagnostic classifiers based on patient gene expression.

From: Upper airway gene expression reveals suppressed immune responses to SARS-CoV-2 compared with other respiratory viruses

Fig. 2

a Receiver operating characteristic (ROC) curve for a 27-gene classifier that differentiates COVID-19 from other acute respiratory illnesses (viral and non-viral). The mean and range of the area under the curve (AUC) are indicated. b Accuracy of the 27-gene classifier within each patient group using a cut-off of 40% out-of-fold predicted probability for COVID-19. c ROC curve for a 10-gene classifier. d ROC curve for a 3-gene classifier. e Out-of-fold predicted probability of COVID-19 derived from the 27-gene classifier plotted as a function of SARS-CoV-2 viral load, log10(rpM). Dashed lines indicate 40% (our chosen cut-off) and 50%.

Back to article page