Fig. 7: Blind study with clinical samples comparing PON1-HDL to other biomarkers from different platforms. | Nature Communications

Fig. 7: Blind study with clinical samples comparing PON1-HDL to other biomarkers from different platforms.

From: Quantifying PON1 on HDL with nanoparticle-gated electrokinetic membrane sensor for accurate cardiovascular risk assessment

Fig. 7

a PON1-HDL using NGEMS (AUC = 0.99). b PON1-HDL using ELISA-1 (AUC = 0.90). c Total PON1 using sandwich ELISA (AUC = 0.83). d PON1 activity assay using rate of reaction with substrate (AUC = 0.67). e Non-HDL cholesterol (AUC = 0.645), f HDL cholesterol (AUC = 0.64) and g cholesterol ratio (AUC = 0.65) from cholesterol assay. h Triglyceride levels from triglyceride assay (AUC = 0.65). Total HDL particle concentration HDL-P from i competitive ELISA (AUC = 0.62), j ELISA-2 (AUC = 0.775), k NGEMS as sum of PON1-HDL and PON1-free HDL (AUC = 0.645) and l 1H-NMR (AUC = 0.61). m PON1-free HDL from NGEMS (AUC = 0.52). n ApoAI from sandwich ELISA (AUC = 0.69). o ApoB from sandwich ELISA (AUC = 0.60). p OxLDL from commercial ELISA (AUC = 0.67). p-values are calculated from unpaired parametric one-tailed t-test with Welch’s correction. AUC values calculated from ROC plots (see Supplementary Fig. 3 for ROC plots). Each datapoint was average of several replicates. Every sample in the figures are measured in triplicates except in e–g where they are duplicated. Each plot is a standard box and whisker plot with central line being the median, box being the 25th and 75th quantile and whiskers representing 0th and 100th quantile. Both controls and CAD groups had ten samples (n = 10) each.

Back to article page