Partially sequenced organisms, decoy searches and false discovery rates

Area of Science:

• Proteomics
• Bioinformatics
• Mass Spectrometry

Background:

• Tandem mass spectrometry (MS/MS) is crucial for peptide identification by matching experimental spectra to database-predicted spectra.
• False discovery rate (FDR) is the standard metric for assessing the quality of peptide identifications.
• Current FDR estimation methods are primarily applied to completely sequenced organisms or known mixtures.

Purpose of the Study:

• To investigate the applicability of FDR estimation for peptide identification in partially sequenced organisms.
• To evaluate the robustness of the PeptideProphet mixture model approach with incomplete sequence databases.

Main Methods:

• Utilized real human plasma data and simulated partial sequence databases.
• Employed the PeptideProphet mixture model for FDR estimation.
• Incorporated decoy sequences to enhance identification accuracy.

Main Results:

• Demonstrated that the PeptideProphet mixture model is robust for partial sequence databases.
• Showed that combining the model with decoy sequences improves FDR estimation reliability.
• Confirmed the applicability of FDR estimation even when peptides are absent from the searched database.

Conclusions:

• Recommends using the PeptideProphet mixture model with decoy sequences for FDR estimation in incompletely sequenced organisms.
• Highlights the importance of robust quality metrics for peptide identification in proteomic studies.
• Suggests this approach enhances the reliability of reporting peptide identifications from diverse biological samples.