ProteomeGenerator: A framework for comprehensive proteomics based on de novo transcriptome assembly and high-accuracy peptide mass spectral matching

Abstract

SUMMARYModern mass spectrometry now permits genome-scale and quantitative measurements of biological proteomes. However, analyses of specific specimens are currently hindered by the incomplete representation of biological variability of protein sequences in canonical reference proteomes, and the technical demands for their construction. Here, we report ProteomeGenerator, a framework for de novo and reference-assisted proteogenomic database construction and analysis based on sample-specific transcriptome sequencing and high-resolution and high-accuracy mass spectrometry proteomics. This enables assembly of proteomes encoded by actively transcribed genes, including sample-specific protein isoforms resulting from non-canonical mRNA transcription, splicing, or editing. To improve the accuracy of protein isoform identification in non-canonical proteomes, ProteomeGenerator relies on statistical target-decoy database matching augmented with spectral-match calibrated sample-specific controls. We applied this method for the proteogenomic discovery of splicing factor SRSF2-mutant leukemia cells, demonstrating high-confidence identification of non-canonical protein isoforms arising from alternative transcriptional start sites, intron retention, and cryptic exon splicing, as well as improved accuracy of genome-scale proteome discovery. Additionally, we report proteogenomic performance metrics for the current state-of-the-art implementations of SEQUEST HT, Proteome Discoverer, MaxQuant, Byonic, and PEAKS mass spectral analysis algorithms. Finally, ProteomeGenerator is implemented as a Snakemake workflow, enabling open, scalable, and facile discovery of sample-specific, non-canonical and neomorphic biological proteomes (https://github.com/jtpoirier/proteomegenerator).

Splice-Junction-Based Mapping of Alternative Isoforms in the Human Proteome

“…The generation of accurate protein sequence databases is an important step in avoiding inflation of false positives during database search and entails finding the set of isoform peptides that exists in a particular sample and is detectable by the MS experimental design. Recent studies have used highthroughput sequencing reads as a template to identify variant protein sequences (Cifani et al, 2018;Carlyle et al, 2018;Zickmann and Renard, 2015;Verbruggen et al, 2019;Mertins et al, 2016;Wang et al, 2019). Our approach builds on prior work and is distinguished by the selection for splice junction pairs in AS events with appreciable RNA-seq read counts.…”

“…Several approaches have been proposed to improve MS identification of AS isoforms, including the curation of splice variant databases (Tavares et al, 2014;Mo et al, 2008) and de novo 6-frame translation of genome sequences (Power et al, 2009;Fermin et al, 2006). More recently, RNA-seq has been leveraged with some success to identify variant sequences not found in standard protein databases (Ning and Nesvizhskii, 2010;Zickmann and Renard, 2015;Verbruggen et al, 2019;Cifani et al, 2018), corroborating the potential utility of an RNAguided approach for discovering protein AS isoforms. Thus far, however, studies of this type have largely been performed in transformed cell lines or tumors known to have aberrant splicing (Ning and Nesvizhskii, 2010;Koch et al, 2014;Sheynkman et al, 2013;Evans et al, 2012;Liu et al, 2017).…”

Discovery of Protein Modifications Using Differential Tandem Mass Spectrometry Proteomics

“…We chose to study the response of mouse RAW264.7 macrophage cells to lipopolysaccharide (LPS), a potent inducer of macrophage activation and differentiation that involves extensive protein and metabolic signaling (Alasoo et al 2015;Kamal et al 2018;Rattigan et al 2018;Seim et al 2019). We treated RAW264.7 cells with LPS using established methods and analyzed the resultant proteomes by two-dimensional nanoscale liquid chromatography and high-resolution mass spectrometry proteomics (Cifani et al 2018). Using SAMPEI parameters designed to maximize the specificity of unbiased discovery (MPI ≥ 0.5, LGP  0.4), we identified 21,846 unique PSMs with putative peptide modifications producing mass shift alignments between 10 and 200 Da (Figure 3b, Supplementary Figure 4).…”

Exploiting Interdata Relationships in Next-generation Proteomics Analysis

“…Another factor lowering RNA and protein correlations arises from alternative splicing and the production of protein isoforms, and many efforts in integrative proteomics concern the complete mapping of the resulting proteome diversity (13, 15, 23–25). Integrating data from mRNA sequencing and proteomics, Liu et al attempted to map the entire human proteome with respect to its variants (26).…”

“…Other examples arise from the inclusion of sequencing data that identifies ribosome footprint positions along mRNAs, estimating translation efficiency and regulatory elements (30). Several such studies exist and examined meiosis or the response to environmental stress (13, 15, 23–25, 31, 32). For example, when comparing genome-wide transcriptome, proteome, and ribosome profiles across diverse stresses, Ho et al found that ribosomes appeared to dissociate from some transcripts, delaying their translation into the corresponding proteins (32).…”