Peptide-level Robust Ridge Regression Improves Estimation, Sensitivity, and Specificity in Data-dependent Quantitative Label-free Shotgun Proteomics

Abstract

Peptide intensities from mass spectra are increasingly used for relative quantitation of proteins in complex samples. However, numerous issues inherent to the mass spectrometry workflow turn quantitative proteomic data analysis into a crucial challenge. We and others have shown that modeling at the peptide level outperforms classical summarization-based approaches, which typically also discard a lot of proteins at the data preprocessing step. Peptide-based linear regression models, however, still suffer from unbalanced datasets due to missing peptide intensities, outlying peptide intensities and overfitting. Here, we further improve upon peptide-based models by three modular extensions: ridge regression, improved variance estimation by borrowing information across proteins with empirical Bayes and M-estimation with Huber weights. We illustrate our method on the CPTAC spike-in study and on a study comparing wild-type and ArgP knock-out Francisella tularensis proteomes. We show that the fold change estimates of our robust approach are more precise and more accurate than those from stateof-the-art summarization-based methods and peptidebased regression models, which leads to an improved sensitivity and specificity. We also demonstrate that ionization competition effects come already into play at very low spike-in concentrations and confirm that analyses with peptide-based regression methods on peptide intensity values aggregated by charge state and modification status (e.g. High-throughput LC-MS-based proteomic workflows are widely used to quantify differential protein abundance between samples. Relative protein quantification can be achieved by stable isotope labeling workflows such as metabolic (1, 2) and postmetabolic labeling (3-6). These types of experiments generally avoid run-to-run differences in the measured peptide (and thus protein) content by pooling and analyzing differentially labeled samples in a single run. Labelfree quantitative (LFQ) 1 workflows become increasingly popular as the often expensive and time-consuming labeling protocols are omitted. Moreover, LFQ proteomics allows for more flexibility in comparing samples and tends to cover a larger area of the proteome at a higher dynamic range (7,8). Nevertheless, the nature of the LFQ protocol makes shotgun proteomic data analysis a challenging task. Missing values are omnipresent in proteomic data generated by data-dependent acquisition workflows, for instance because of low-abundant peptides that are not always fragmented in complex peptide mixtures and a limited number of modifications and mutations that can be accounted for in the feature search. Moreover, the overall abundance of a peptide is determined by the surroundings of its corresponding cleavage sites as these influence protease cleavage efficiency (9). Similarly, some peptides are more easily ionized than others (10). These issues not only lead to missing peptides, but also increase variability in individual peptide intensities. The discrete nature of MS1 sampling following continuous ...

MSqRob takes the missing hurdle: uniting intensity- and count-based proteomics

“…MSqRob has been described in detail in Goeminne et al (2016) 9 . Briefly, for each protein, the log2-transformed peptide intensities for peptide = 1, … , in run = 1, … , are modeled as follows:…”

Molecules and elements for quantitative bioanalysis: The allure of using electrospray, MALDI, and ICP mass spectrometry side‐by‐side

“…Generally, for quantification, either stable isotopic labels are introduced in the analytes followed by isotope dilution analysis (IDA) or chromatographic signal intensities are correlated with peptide or protein abundances. The latter way called “label free” avoids chemical reactions, but has limitations in accuracy and has per se no multiplexing options …”

“…The latter way called "label free" avoids chemical reactions, but has limitations in accuracy and has per se no multiplexing options. [6][7][8][9][10] Generally, the alternative methodology relies on labeling peptides with stable isotopes introducing a well-defined and sufficient mass difference between samples and standard. The labeling can be done in many ways and stable isotopes are most commonly 15 N and 13 C or 18 O and many possible combinations of them, which allow using MS/ MS methods or very high resolution MS or combinations of both.…”

BayesENproteomics: Bayesian elastic nets for quantification of proteoforms in complex samples

“…The effects of outlier peptides on the final linear model fit can be mediated by various weighting strategies (e.g. see Goeminne et al (2016)). However, for modelling changes in PTM abundance, it would be useful to obtain effect sizes for peptides that specifically interact with a given treatment and separate them from effect sizes for other parameters that may be interacting with these peptides.…”

“…Simpler models (e.g. those used by (Choi et al (2014) and (Goeminne et al (2016)) minimise the potential for overfitting found in more complex models. However, when many potential sources of variability exist it may be necessary to fit increasingly complex models wherein the choice of terms included, particularly any interaction terms, may need to be decided on a protein-by-protein basis.…”