Skip to main content


Fig. 4 | Journal of Translational Medicine

Fig. 4

From: Decreased plasma phospholipid concentrations and increased acid sphingomyelinase activity are accurate biomarkers for community-acquired pneumonia

Fig. 4

Identification of diagnostic biomarkers by ROC curve analysis. Differential expression (ratio of medians, also referred to as “fold change”; y-axis) is plotted against area under the ROC curve (AUC, x-axis), with darkness of fill color indicating the asymptotic significance of the ROC curve. Each dot corresponds to one lipid analyte (ac) or one metabolite indicator (df). The dotted vertical line highlights the cut-off for “excellent classification” [26], i.e. AUC 0.80. a, d CAP baseline vs. Controls; b, e COPD vs. Controls; c, f CAP baseline vs. COPD The ROC analysis parameters of all markers with AUC ≥ 0.8 for each comparison are listed in Additional file 1: Table S2. gl Levels of the most accurate biomarker for each of the three paired comparisons shown in af. gi Best lipid biomarkers; PCaaC34:4 (CAP vs. CTRL), PCaaC42:6 (COPD vs. Control), lysoPCaC20:4 (CAP vs. COPD). jl Best metabolic indicator biomarkers; Total lysoPC (CAP vs. Control), Total SM/Total PC (COPD vs Control), total SM (CAP vs. COPD). The box spans the 25th–75th percentile; lower whiskers = bottom quartile; upper whisker = upper quartile; and horizontal line = median. *p < 0.05, **p < 0.01, ***p < 0.001 (Mann–Whitney U test). Metabolite: lysoPC, lysophosphatidylcholines; PC, phosphatidylcholines; SM, sphingomyelins

Back to article page