pFind Studio: a computational solution for mass spectrometry-based proteomics

Introduction

Database search has long been the dominant approach to peptide and protein identification. Many database search algorithms are used in the routine proteome analysis, such as pFind, Mascot, SEQUEST, PEAKS DB. Generally, the essence of these methods is retrieving all candidate peptides from a specified database for each spectrum, which also means that a protein database is indispensable.

An alternative method is de novo peptide sequencing, which deduces peptide sequences directly from MS/MS data without any databases. De novo peptide sequencing can identify mutations, unexpected modifications and novel peptides. pNovo is a de novo peptide sequencing tool, which has been updated with multiple versions ( e.g., pNovo and pNovo+).

Figure 1. Workflow of Open-pNovo. (a) Example of an original spectrum from a peptide AN(Met[H])VR, where Met[H] denotes the methylation of Histidine residue. (b) Spectrum graph (DAG) for the original spectrum. The black edges denote the normal edges, while the red ones denote the modification ones. If we do not consider the red edge, the correct peptide cannot be obtained. (c) Normal peptides and the modified peptides are obtained by finding k longest paths in DAG. (d) Score function is trained by RankBoost using three features. (e) Candidate peptides are output and sorted by the final scores, which are obtained by RankBoost. Note that when translating the original spectrum a to the spectrum graph b, each peak is translated into two vertices rather than one vertex in panel b

®

Copyright © pFind Team · Institute of Computing Technology, Chinese Academy of Sciences, Beijing, China