Evaluating de novo sequencing in proteomics: already an accurate alternative to database-driven peptide...

Area of Science:

• Proteomics
• Computational Biology
• Mass Spectrometry

Background:

• Database search methods dominate mass spectrometry (MS)-based shotgun proteomics.
• High-resolution MS data enables improved computational de novo peptide sequencing.
• De novo sequencing bypasses the need for reference databases.

Purpose of the Study:

• To evaluate the performance of de novo sequencing algorithms on high-resolution MS data.
• To compare the accuracy and speed of Novor, PEAKS, and PepNovo.
• To identify common de novo sequencing errors and influencing factors.

Main Methods:

• Processed four experimental MS datasets from different instrument types and fragmentation modes (collision-induced dissociation and higher energy collisional dissociation).
• Evaluated Novor, PEAKS, and PepNovo software packages.
• Tested algorithm accuracy on simulated spectra generated from peak intensity prediction software.

Main Results:

• Novor exhibited the best performance in predicting full peptide, tag-based, and single-residue sequences.
• Novor achieved a 12-17x speedup compared to PEAKS.
• Algorithms showed moderate accuracy (approx. 35%) on experimental data but significantly higher accuracy (up to 84%) on simulated data.

Conclusions:

• Novor is the top-performing algorithm for de novo peptide sequencing in terms of accuracy and speed.
• De novo sequencing shows promise for broader application in proteomics, especially with advancements in data quality and algorithms.
• Understanding sequencing errors and data quality factors is crucial for improving de novo sequencing reliability.