(A) ROC curve and AUC for the prediction model of COVID-19 severity, i.e., critical COVID-19 (n = 50) or mild COVID-19 (n = 35) obtained by LASSO logistic regression. Included predictors were age and mRNA expression of MUC16, MUC20, and MUC21, with an AUC of 89.1% (95% CI: 80.4%–96.1%), a sensitivity of 90.0% (95% CI: 79.3%–96.2%), and a specificity of 85.7% (95% CI: 71.2%–94.6%). (B) ROC curve and AUC for the prediction model of COVID-19 presentation: COVID-19 positive (n = 85) or COVID-19 negative (n = 30) obtained by the LASSO logistic regression. Included predictors were age and mRNA expression of MUC1, MUC2, MUC4, MUC6, MUC13, MUC16, and MUC20, with an AUC of 91.8% (95% CI: 85.2%–97.1%), a sensitivity of 90.6% (95% CI: 83.0%–95.5%), and a specificity of 93.3% (95% CI: 79.7%–98.9%).