Skip to main content
. 2017 Oct 15;18(10):2136. doi: 10.3390/ijms18102136

Figure 1.

Figure 1

CTCL-MF patients and miR-classifier for CTCL plasma. (A) Photographs of the lesions in MF patients (MF1-7). Each panel shows a larger view of the lesion and its detail; (B) Principal component analysis (PCA) analysis of the miR-classifier in CTCL, benign lesions (BL) and healthy donors using expression data of miR-155, miR-203, miR-205, miR-22, miR-223, miR-103, and miR423-5p. A line separating the CTCL from BL is defined in the Results section. MF: dark triangle, SS: grey rectangle, BL: empty triangle, and controls are indicated by a cross. The arrows indicate two CR samples of one patient. (B) Right: classification performance in the training set using the progressive sampling algorithm called NSC (specificity 100%, sensitivity 94%) in which the dark field represents samples correctly diagnosed as malignant, the grey field represents samples correctly diagnosed as benign, and empty fields represent samples that are falsely diagnosed. Dark person icon: CTCL; grey person icon: BL.