Fig. 6

Workflow using the best retention time predictor (SSRCalc in RTCalc) to filter a list of PSMs based on the agreement between measured and predicted retention time, assuming the identification is correct (a). The user selects a Z-score threshold to remove outliers, which are likely due to false identifications. Here we used a maximum absolute Z-score of 2 to demonstrate the workflow, although this may be overly conservative, as a few PSMs of very high PeptideProphet probability are also removed (b). The workflow is available on myExperiment (workflow #4042)