The Peptide Validator node validates PSMs and peptides in the consensus workflow. This node sets the confidence levels for the PSMs and peptides by using PEPs, q-values, and false discovery rate (FDR) calculations.

You can use four different approaches to validate the results: PSM-level false discovery rate (FDR) calculation based on search-engine-specific main scores, controlling the peptide-level error rate, automatically controlling the peptide-level error rate, or not controlling the peptide-level error rate. Use the Validation Mode parameter of the Peptide Validator node to specify one of these settings.

  • Automatic (Control Peptide-Level Error Rate If Possible): First, the application determines whether Percolator-generated q-values and PEPs are available for PSMs. If they are, it uses them to assign the PSM confidences (high, medium, or low) according to the settings of the user-defined Target FDR (Strict) for PSMs parameter and the Target FDR (Relaxed) for PSMs parameter. Next, the application uses the qvality algorithm to calculate q-values and PEPs for the peptides. Finally, it assigns peptide confidences (high, medium, or low) according to the settings of the user-defined Target FDR (Strict) for Peptides parameter and the Target FDR (Relaxed) for Peptides parameter.
  • If no Percolator-generated q-values and PEPs are available for PSMs, the application sets the PSM confidences according to the user-defined Target FDR (Strict) for PSMs parameter and the Target FDR (Relaxed) for PSMs parameter, using their respective search engine scores (for example, XCorr vs. charge state for Sequest HT). The peptide confidences are then represented by the confidence of their respective best PSM.
  • Only PSM-Level FDR Calculation Based on Score: The application does not use any PEPs and q-values, even if they are available, to assign PSM confidences. It bases PSM confidences on the user-defined Target FDR (Strict) for PSMs parameter and the Target FDR (Relaxed) for PSMs parameter, using their respective search engine scores (for example, XCorr vs. Charge State for Sequest HT). The peptide confidences are then represented by the confidence of their respective best PSM.
  • Control Peptide-Level Error Rate (Calculate Missing q-values for PSMs): First, the application determines whether Percolator-generated q-values and PEPs are available for PSMs. If they are, it uses them to assign the PSM confidences (high, medium, or low) according to the user-defined Target FDR (Strict) for PSMs parameter and the Target FDR (Relaxed) for PSMs parameter.
  • If no Percolator-generated q-values and PEPs are available for the PSMs, the application tries to calculate those values by running quality on the main search engine score.
  • Next, the application uses the quality algorithm to calculate q-values and PEPs for peptides. It then assigns peptide confidences (high, medium, or low) according to the settings of the user-defined Target FDR (Strict) for Peptides parameter and the Target FDR (Relaxed) for Peptides parameter.
  • If the application cannot obtain q-values and PEPs for the PSMs by using quality, it sets the PSM confidences according to the user-defined Target FDR (Strict) for PSMs parameter and the Target FDR (Relaxed) for PSMs parameter, using their respective search engine scores (for example, XCorr vs. charge state for Sequest HT). The peptide confidences are then represented by the confidence of their respective best PSM.
  • Automatic (Without Control of Peptide-Level Error Rate): First, the application determines whether Percolator-generated q-values and PEPs are available for PSMs. If they are, it uses them and assigns the PSM confidences (high, medium, or low) according to the settings of the user-defined Target FDR (Strict) for PSMs parameter and the Target FDR (Relaxed) for PSMs parameter.
  • If no Percolator-generated q-values and PEPs are available for PSMs, the application sets the PSM confidences according to the user-defined Target FDR (Strict) for PSMs parameter and the Target FDR (Relaxed) for PSMs parameter, using their respective search engine scores (for example, XCorr vs. charge state for Sequest HT).
  • In both cases, though, the peptide confidences are represented by the confidence of their respective best PSM. In other words, the application does not use the quality algorithm to calculate peptide confidences, so it does not control the error rate on the peptide level.
  • This particular validation mode is especially useful when you combine and compare results where PEPs and q-values are available only for a subset of the data.

 

Connect the Peptide Validator node to the PSM Grouper node and the Peptide and Protein Filter node in the consensus workflow.

The following table describes the parameters for the Peptide Validator node.

Peptide Validator node parameters

Parameter

Definition

Validation Mode

Specifies the mode used to validate identification results.

  • (Default) Automatic (Control Peptide-Level Error Rate if Possible)
  • Automatic (Without Control of Peptide-Level Error Rate)
  • Only PSM-Level FDR Calculation Based on Score
  • Control Peptide-Level Error Rate (Calculate Missing q-values for PSMs)

Target FDR (Strict) for PSMs

Specifies the target false discovery rate for PSM matches of high confidence.

Range: 0.0–1.0; default: 0.01

Target FDR (Relaxed) for PSMs

Specifies the target false discovery rate for PSM matches of medium confidence.

Range: 0.0–1.0; default: 0.05

Target FDR (Strict) for Peptides

Specifies the target false discovery rate for peptide matches of high confidence.

Range: 0.0–1.0; default: 0.01

Target FDR (Relaxed) for Peptides

Specifies the target false discovery rate for peptide matches of medium confidence.

Range: 0.0–1.0; default: 0.05

Validation Based On

Determines the value that the validation is based on.

  • PEP: Bases the PSM validation on PEP values.
  • (Default) q-value: Bases the PSM validation on q-values.

Target/Decoy Selection for PSM Level FDR Calculation Based on Score

Selects Target/Decoy Strategy (Separate/Concatenated) for PSM Level FDR based on score. In automatic mode, the software checks whether all searches were validated in same mode in the processing step. If the validation mode is the same, it is also applied in this node, otherwise the concatenated mode is used. Automatic is default.

 

Reset Confidences for Nodes Without Decoy Search (Fixed Score Thresholds)

Sets confidence levels for nodes that do not conduct a decoy search on the basis of current server settings (thresholds) for the specific search type.

  • True: Resets the confidence levels by using the thresholds set on the view opened by the Administration > Configuration command for the different search node types.
  • (Default) False: Retains the PSM confidence levels from searches without decoy results.