Skip to main content
. 2021 Mar 8;12:636289. doi: 10.3389/fimmu.2021.636289

Figure 6.

Figure 6

(A) AUROC and PR curves for classifiers predicting whether a gene expression sample is from a Lyme disease diagnosed patient or a healthy control across all visits. (B) AUROC And PR curves for classifiers predicting whether a gene expression sample is from a Lyme disease diagnosed patient with persistent symptoms. (C) AUROC and PR curves for classifiers predicting whether a gene expression sample is from a Lyme disease diagnosed patient or a healthy control across all visits using CIBERSORT features instead of RNA-seq gene expression. (D) AUROC and PR curves for classifiers predicting whether a gene expression sample is from a Lyme disease diagnosed patient with persistent symptoms using CIBERSORT features instead of RNA-seq gene expression. (E) AUROC and PR curves for classifiers predicting whether a patient's observed gene expression, normalized and batch effect corrected across experiments, suggests they have Lyme disease or COVID-19.