Search bioRxiv⌕ Search

Biology subjects

Zumer, J.

Publications and source records attributed to Zumer, J..

2 recordsLinked to original sources
bioRxiv2023

Pepid: a Highly Modifiable, Bioinformatics-Oriented Peptide Search Engine

MotivationCurrent peptide search engines are optimized for wet-lab workflows, i.e. they operate in an "end-to-end" manner to achieve good identification results, not to be modified or provide algorithmic insight. This makes developing new software methods to solve problems in peptide identification methods difficult, often requiring a full engine rewrite. Recently, many deep learning methods were proposed as solutions to various parts of the peptide identification task, but virtually none of those methods have been implemented in any actual peptide search process. We believe that the lack of a reliable bioinformatics research platform for peptide identification that enables such integrations is slowing down proteomics research as a whole. ResultsWe present pepid, a bioinformatics research-oriented peptide search engine. Unlike other search engines, pepid is specifically designed with ease of computational research in mind. Our design is highly flexible and allows easy modifications with little required software development expertise, allowing researchers to focus on analysing and improving peptide identification methods.It also takes recent computational trends into account, such as the recent slew of deep learning publications in proteomics, and features a multi-phased batched operations design that is more appropriate than the spectrum batch "end-to-end" designs of existing search engines for those approaches. We show that pepid is competitive with common engines in terms of both identification rates and runtime, forming a minimum required baseline to enable further identification research. Availability and ImplementationPepid is available as open source software under the MIT license at https://github.com/lemieux-lab/pepid. Other data referenced in the text is 3rd party. The selected yeast proteome can be found on SwissProt with accession ID UP000002311 while the human proteomes accession ID is UP0000005640. The ProteomeTools spectra can be found in the PRIDE archive under accession D PXD004732 and the One Hour Yeast Proteome can be found at the ChorusProject at https://chorusproject.org/anonymous/download/experiment/-8823069691100997209 and https://chorusproject.org/anonymous/download/experiment/449795368199176159.

bioinformatics↗
bioRxiv2023

Mining Mass Spectra for Peptide Facts

The current mainstream software for peptide-centric tandem mass spectrometry data analysis can be categorized as either database-driven, which rely on a library of mass spectra to identify the peptide associated with novel query spectra, or de novo sequencing-based, which aim to find the entire peptide sequence by relying only on the query mass spectrum. While the first paradigm currently produces state-of-the-art results in peptide identification tasks, it does not inherently make use of information present in the query mass spectrum itself to refine identifications. Meanwhile, de novo approaches attempt to solve a complex problem in one go, without any search space constraints in the general case, leading to comparatively poor results. In this paper, we decompose the de novo problem into putatively easier subproblems, and we show that peptide identification rates of database-driven methods may be improved in terms of peptide identification rate by solving one such subsproblem without requiring a solution for the complete de novo task. We demonstrate this using a de novo peptide length prediction task as the chosen subproblem. As a first prototype, we show that a deep learning-based length prediction model increases peptide identification rates in the ProteomeTools dataset as part of an Pepid-based identification pipeline. Using the predicted information to better rank the candidates, we show that combining ideas from the two paradigms produces clear benefits in this setting. We propose that the next generation of peptide-centric tandem mass spectrometry identification methods should combine elements of these paradigms by mining facts "de novo; about the peptide represented in a spectrum, while simultaneously limiting the search space with a peptide candidates database.

bioinformatics↗