Skip to main content
Fig. 5 | Journal of Translational Medicine

Fig. 5

From: Prediction of occult tumor progression via platelet RNAs in a mouse melanoma model: a potential new platform for early detection of cancer

Fig. 5

ROC curves of the diagnostic performances of gene predictors for occult tumor progression in mice. ROC curves for the diagnostic performances of the prediction score formulas generated from LASSO regression in the mouse cohort. ROC curves for the discrimination of early-early tumor (occult tumor, E group) from negative control group (C group) (a, E vs. C) or from macroscopic melanoma group (M group) (b, E vs. M). Probability statistics calculated according to the prediction score formulas generated from LASSO regression analyses: Probability = eScore / (1 + eScore). 95% CI of AUC: training data 0.872–0.999, testing data 0.799–1.000 (a); training data 0.852–0.988, testing data 0.837–1.000 (b)

Back to article page