Table 2.
Confusion matrices and performance metrics at different proteolytic contamination count scores
| Peptide subsets and performance metrics | PCC ≥ 0 | PCC ≥ 1 | PCC ≥ 2 | PCC ≥ 3 |
|---|---|---|---|---|
| Contamination marked as contamination (True Positive) | 84 | 72 | 43 | 10 |
| Ligand marked as contamination (False Positive) | 416 | 28 | 2 | 0 |
| Ligand marked as ligand (True negative) | 0 | 388 | 414 | 416 |
| Contamination marked as ligand (False negative) | 0 | 12 | 41 | 74 |
| Percentage of detected contaminations (Sensitivity) | 100.00% | 85.70% | 51.20% | 11.90% |
| Percentage of detected ligands (Specificity) | 0.00% | 93.30% | 99.50% | 100.00% |
Table includes confusion matrices showing the number of peptides in each group (true/false positive and true/false negatives) as well as performance metrics (sensitivity and specificity) for contamination assignment at different proteolytic contamination count (PCC) scores using expert review benchmark EC500.