Fig. 4

Confidence interval: Confidence interval for sensitivity and specificity across all machine learning algorithms, namely K-nearest neighbors (KNN), Naïve Bayes (NB), neural network (NNET), and random forest (RF). Known biomarker (A: all transcripts, B: protein coding transcripts, and C: non-coding transcripts) and all data (D: all transcripts, E: protein coding transcripts, and F: non-coding transcripts). For all the calculations, transcript expression was used