Skip to main content
Fig. 4 | Clinical Proteomics

Fig. 4

From: An Unsupervised, Model-Free, Machine-Learning Combiner for Peptide Identifications from Tandem Mass Spectra

Fig. 4

Left column: ROC curves for data sets S17, AURUM, and OMICS—C-TMO (dash-dotted line), heuristic combiner (dashed line), U-TMO (solid line), and U*-TMO (dotted line) shown. Right column: AUROC after each training iteration for data sets S17, AURUM, and OMICS—true protein labels at every iteration (dash-dotted and asterisks), true protein labels for initialization only (solid and boxes), heuristic protein labels for initialization (dashed and circles), and single best consensus protein labels for initialization (dotted and crosses)

Back to article page

Clinical Proteomics

ISSN: 1559-0275