PeptideDetective™ de novo Peptide Sequencing Software
"Make the Right Call!℠"

The Challenge:

Correctly identifying a peptide from its MS/MS spectrum is an often daunting and deceptively misleading task. Traditional database-matching strategies for sequencing miss novel sequences that are not in the database, while de novo sequencing approaches are severely challenged to select a reliable result from the vast combinatorial possibilities for candidates. The de novo approaches can quickly become computationally demanding, and are generally much less robust or reliable than the database approaches.

Even with "ideal" data (presentation link below), current top-performing de novo software incorrectly identifies data more than 15% of the time! Such mis-identifications waste valuable time and costs in avoiding or recovering from misdirected R&D efforts.

The de novo challenge is far greater with real-world data. Results are highly vulnerable to the mass tolerance(s) selected by the user (presentation link below). Optimal performance requires choosing a mass tolerance that strikes the ideal balance between being large enough to encompass the true sequence identity but small enough to exclude the landslide of incorrect competing sequences that can quickly overwhelm de novo techniques.

The PeptideDetective™ Solution:

PeptideDetective software helps you substantially reduce and avoid incorrect calls and misdirected R&D. It leverages information theory and high-performance computing to substantially improve peptide sequencing accuracy beyond the limits of current leading software offerings. In a software competition benchmarking PeptideDetective against PepNovo+ (Digital Proteomics) and PEAKS® 7 (Bioinformatics Solutions Inc.), on 168 synthesized known peptides, these current software solutions performed 10-81% worse than PeptideDetective on correct identifications in the Top 1, 5, or 10 candidates offered by each competitor.

Of particular note, current software performance typically varies dramatically across different user parameter settings for mass tolerance. Users are challenged to provide a best estimate of mass tolerance conditions for spectra "as acquired" - not so tight that they exclude the correct sequence from consideration, while not so wide that they lose the answer in the growing deluge of false positive candidates. PeptideDetective simply requires a single mass tolerance setting that is wide enough to ensure that correct solution ions are included in the dataset, and otherwise is robust to wider mass tolerance settings. (The peptide spectra and results of this study are available via the links below.)

Alpha Phase Invitation

As part of our development path towards full commercial release, PeptideDetective is now available for Alpha phase studies performed with our collaborators. The current version supports analysis of singly-charged CID spectra, and multi-charged ion capabilities will follow. Deeper innovations in peptide spectral analysis and PTM support are in development for further extension of the PeptideDetective platform.

We seek a variety of spectra to analyze from different MS instruments and different identification challenges. Typically, we seek to perform and share a comparison of results versus current software offerings, where results are independently supported or verified, and results may be shared in scientific meetings and publications.

We also engage with partners providing existing software solutions, in order to provide new added-value to smartly extend and reinforce the capabilities of their existing offerings.

To explore potential Alpha phase collaborations utilizing PeptideDetective, please contact us at info@veritomyx.com or by calling 650-812-8140.

Links

View the Software Comparison Presentation
(PepNovo+ vs. Peaks 7 vs. PeptideDetective)

Download the "Ideal" Control Data used in the Software Comparison Presentation
(recalibrated to known identities within)

Download the Realistic User Data used in the Software Comparison Presentation
(presented on "as acquired" basis)

Posters

“Optimizing Collision Energy in Collision-Induced Dissociation for Peptide Sequencing" (US HUPO 2018)


Abstract:
Although there have been advances in mass spectrometers, ionization techniques and hybrid data collection approaches, it remains a challenge to derive a peptide’s sequence from its product ion spectrum. Recent bottom up proteomic proficiency studies1,2 still show low sequencing efficiency and high error rates on known peptides. At least part of the problem may be related to collision-induced dissociation (CID) conditions. Previous studies3,4 show that CID parameters have a pronounced effect on the fragmentation of a given peptide ion and determine the useful sequencing information found in the spectrum. The goal of this study is to examine how collision energy (CE) affects the information content of the resulting MS2 spectrum, and how optimal CID conditions may change with peptide composition.


Citation:
Sokkalingam N, Schneider L, Wright W, Ashrafi S, Tenderholt A, Peterson J, Duncan M. Optimizing Collision Energy in Collision-Induced Dissociation for Peptide Sequencing. Poster presented at: 14th Annual US HUPO Conference; 2018 Mar 11-14, Minneapolis, MN, USA.


[Download PDF of poster]