Figure 1
From: Biomarker selection and a prospective metabolite-based machine learning diagnostic for lyme disease
(a) UMAP visualization of log transformed and KNN imputed LC-MS data from training samples. EDL early disseminated Lyme disease, ELL early localized Lyme disease, HCN healthy control non-endemic, HCE1 healthy control endemic site 1. (b) UMAP visualization of log transformed and KNN imputed LC-MS data from training samples post IFR. EDL early disseminated Lyme disease, ELL early localized Lyme disease, HCN healthy control non-endemic, HCE1 healthy control endemic site 1. (c) UMAP visualization of log transformed and KNN imputed LC-MS data from training samples restricted to the features found by IFR. EDL early disseminated Lyme disease, ELL early localized Lyme disease, HCN healthy control non-endemic, HCE1 healthy control endemic site 1.
