Skip to main content

Advertisement

Table 3 Class Prediction Model Results.

From: Proteomic profiling of urine for the detection of colon cancer

Dataset Sensitivity Specificity Misclassification Rate AUROC
MALDI 65 (52–78) 84 (75–93) 24 0.80
MALDI-DS 68 (55–80) 68 (57–79) 32 0.67
SELDI 51 (38–64) 75 (64–86) 34 0.69
ALL 78 (66–90) 87 (78–96) 17 0.88
  1. Table 3 summarises the logistic regression models tested by leave-one-out cross-validation. Sensitivity, specificity and misclassification rate are presented as percentages (95% confidence intervals). The m/z values of the peaks used in the logistic regressions are, MALDI: 1606, 2254, 5011, MALDI-DS: 2193, SELDI: 1885 and 11960, ALL: 1606, 2251, 4758, 5011. Number of samples in each dataset: MALDI: 123 (55 cancer), MALDI-DS: 119 (53 cancer), SELDI: 116 (55 cancer).