Interrogating Fractionation and Other Sources of Variability in Shotgun Proteomes Using Quality Metrics

Abstract

The increasing amount of publicly available proteomics data creates opportunities for data scientists to investigate quality metrics in novel ways. QuaMeter IDFree is used to generate quality metrics from 665 RAW files and 97 WIFF files representing publicly available "shotgun" mass spectrometry datasets. These experiments are selected to represent Mycobacterium tuberculosis lysates, mouse MDSCs, and exosomes derived from human cell lines. Machine learning techniques are demonstrated to detect outliers within experiments and it is shown that quality metrics may be used to distinguish sources of variability among these experiments. In particular, the findings demonstrate that according to nested ANOVA performed on an SDS-PAGE shotgun principal component analysis, runs of fractions from the same gel regions cluster together rather than technical replicates, close temporal proximity, or even biological samples. This indicates that the individual fraction may have had a higher impact on the quality metrics than other factors. In addition, sample type, instrument type, mass analyzer, fragmentation technique, and digestion enzyme are identified as sources of variability. From a quality control perspective, the importance of study design and in particular, the run order, is illustrated in seeking ways to limit the impact of technical variability.

Computational Proteomics: Focus on Deep Learning

“…In the first article, Kriek et al systematically examine the sources of variability in shotgun proteomics by applying machine learning and statistical methods to model quality metrics generated from QuaMeter IDFree. [16] One of the interesting findings is that fractionation is an integral source of variability. In the second article, Eng et al present an updated version of Comet, which enables the search for PTMs and amino acid substitutions encoded in the proteomics standards initiative extended FASTA format (PEFF).…”

“…systematically examine the sources of variability in shotgun proteomics by applying machine learning and statistical methods to model quality metrics generated from QuaMeter IDFree. [ 16 ] One of the interesting findings is that fractionation is an integral source of variability. In the second article, Eng et al.…”

SpectiCal: m/z Calibration of MS2 Peptide Spectra Using Known Low Mass Ions

“…Several software tools − have been developed to analyze both precursor and fragmentation spectra to assess the quality of DDA runs based on information in the spectra. , Some quality control (QC) metrics are computed after a first-round attempt to identify the peptide ions that yielded the fragmentation spectra, and some are computed independently of an identification step, based purely on the spectra themselves (often called “ID free”) . Additionally, it is possible to use peaks in fragmentation spectra for the calibration of instrumental artifacts.…”