Fig. 3: Performance of the stacked ensemble model for cancer detection.
From: Cost-effective methylome sequencing of cell-free DNA for accurately detecting and locating cancer
a ROC curve of the stacked ensemble method for detecting all four cancer types. Source data are provided as a source data file. b Sensitivity breakdown in each cancer stage and cancer type. Sensitivity is shown at 1 false positive (97.9% specificity). The average number of test cancer patients in each cancer type and stage over 10 runs is indicated in the label of the horizontal axis. Sensitivity is not computed if the average number of cancer patients in a cancer stage/type over 10 runs is <4. The points and error bars represent the average sensitivity over 10 runs and 95% confidence intervals. Source data are provided as a Source Data file. c Performance (AUROC) of using all marker types and each individual marker type (\(n:\) 102 samples). The points and error bars represent the average AUROC over 10 runs and 95% confidence intervals. Source data are provided as a Source Data file.
