A decoy-free approach to the identification of peptides.

A growing number of proteogenomics and metaproteomics studies indicate potential limitations of the application of the "decoy" database paradigm used to separate correct peptide identifications from incorrect ones in traditional shotgun proteomics. We therefore propose a binary classifier called Nokoi that allows fast yet reliable decoy-free separation of correct from incorrect peptide-to-spectrum matches (PSMs). Nokoi was trained on a very large collection of heterogeneous data using ranks supplied by the Mascot search engine to label correct and incorrect PSMs. We show that Nokoi outperforms Mascot and achieves a performance very close to that of Percolator at substantially higher processing speeds.

Towards a human proteomics atlas.

Proteomics research has taken up an increasingly important role in life sciences over the past few years. Due to a strong push from publishers and funders alike, the community has also started to freely share its data in earnest, making use of public repositories such as the highly popular PRIDE database at EMBL-EBI. Reuse of these publicly available data has so far been confined to rather specific, targeted reanalyses, but this limited reuse is set to expand dramatically as repositories continue to grow exponentially. Examples of large-scale reuse are readily found in other omics disciplines, where more comprehensive public data have already accumulated over longer periods. Here, a typical example of integrative data reuse is provided by the construction of so-called expression atlases. We here therefore investigate the issues involved in using the human data currently stored in the PRIDE database to construct a robust, tissue-specific protein expression atlas from tandem-MS based label-free quantification.

Identification of Quantitative Proteomic Differences between Mycobacterium tuberculosis Lineages with Altered Virulence.

Evidence currently suggests that as a species Mycobacterium tuberculosis exhibits very little genomic sequence diversity. Despite limited genetic variability, members of the M. tuberculosis complex (MTBC) have been shown to exhibit vast discrepancies in phenotypic presentation in terms of virulence, elicited immune response and transmissibility. Here, we used qualitative and quantitative mass spectrometry tools to investigate the proteomes of seven clinically-relevant mycobacterial strains-four M. tuberculosis strains, M. bovis, M. bovis BCG, and M. avium-that show varying degrees of pathogenicity and virulence, in an effort to rationalize the observed phenotypic differences. Following protein preparation, liquid chromatography mass spectrometry (LC MS/MS) and data capture were carried out using an LTQ Orbitrap Velos. Data analysis was carried out using a novel bioinformatics strategy, which yielded high protein coverage and was based on high confidence peptides. Through this approach, we directly identified a total of 3788 unique M. tuberculosis proteins out of a theoretical proteome of 4023 proteins and identified an average of 3290 unique proteins for each of the MTBC organisms (representing 82% of the theoretical proteomes), as well as 4250 unique M. avium proteins (80% of the theoretical proteome). Data analysis showed that all major classes of proteins are represented in every strain, but that there are significant quantitative differences between strains. Targeted selected reaction monitoring (SRM) assays were used to quantify the observed differential expression of a subset of 23 proteins identified by comparison to gene expression data as being of particular relevance to virulence. This analysis revealed differences in relative protein abundance between strains for proteins which may promote bacterial fitness in the more virulent W. Beijing strain. These differences may contribute to this strain's capacity for surviving within the host and resisting treatment, which has contributed to its rapid spread. Through this approach, we have begun to describe the proteomic portrait of a successful mycobacterial pathogen. Data are available via ProteomeXchange with identifier PXD004165.

An extra dimension in protein tagging by quantifying universal proteotypic peptides using targeted proteomics.

The use of protein tagging to facilitate detailed characterization of target proteins has not only revolutionized cell biology, but also enabled biochemical analysis through efficient recovery of the protein complexes wherein the tagged proteins reside. The endogenous use of these tags for detailed protein characterization is widespread in lower organisms that allow for efficient homologous recombination. With the recent advances in genome engineering, tagging of endogenous proteins is now within reach for most experimental systems, including mammalian cell lines cultures. In this work, we describe the selection of peptides with ideal mass spectrometry characteristics for use in quantification of tagged proteins using targeted proteomics. We mined the proteome of the hyperthermophile Pyrococcus furiosus to obtain two peptides that are unique in the proteomes of all known model organisms (proteotypic) and allow sensitive quantification of target proteins in a complex background. By combining these 'Proteotypic peptides for Quantification by SRM' (PQS peptides) with epitope tags, we demonstrate their use in co-immunoprecipitation experiments upon transfection of protein pairs, or after introduction of these tags in the endogenous proteins through genome engineering. Endogenous protein tagging for absolute quantification provides a powerful extra dimension to protein analysis, allowing the detailed characterization of endogenous proteins.

Structure-based mutant stability predictions on proteins of unknown structure.

The ability to rapidly and accurately predict the effects of mutations on the physicochemical properties of proteins holds tremendous importance in the rational design of modified proteins for various types of industrial, environmental or pharmaceutical applications, as well as in elucidating the genetic background of complex diseases. In many cases, the absence of an experimentally resolved structure represents a major obstacle, since most currently available predictive software crucially depend on it. We investigate here the relevance of combining coarse-grained structure-based stability predictions with a simple comparative modeling procedure. Strikingly, our results show that the use of average to high quality structural models leads to virtually no loss in predictive power compared to the use of experimental structures. Even in the case of low quality models, the decrease in performance is quite limited and this combined approach remains markedly superior to other methods based exclusively on the analysis of sequence features.