Skip to main content
. 2022 Oct 2;23(19):11688. doi: 10.3390/ijms231911688

Figure 5.

Figure 5

Optimization and external validation of the oxylipin signature of MetS. Based on the oxylipins selected in the Discovery and Replication cross-sectional studies (n = 29 oxylipins), a LASSO-penalized conditional logistic regression model including 23 oxylipins was constructed with the participants in the Discovery study. The performances of the LASSO model were assessed in (a) the Discovery and (b) Replication cross-sectional studies using the Area Under the Receiver Operating Characteristic Curve (AUC and error rate were 10-fold cross-validated). Sensitivity is the percentage of Case correctly predicted (“True Positive”) and specificity is the percentage of Control correctly predicted (“True Negative”). TP: True Positive, FP: False Positive, TN: True Negative and FN: False Negative with Positive corresponding to CardMetS and Negative to Control. Circos plot (c) showing the odd ratios (OR) associated with each oxylipin selected in the LASSO model. * represents significant effect of oxylipins (p < 0.05) in the LASSO model. OR are for 1SD-increment of oxylipin absolute concentration. Line colors indicate if OR are >1 (orange) or <1 (blue). Line thickness represent the relative values of OR in the model.