The ability to quantify proteins of interest and ensure the measured quantifications are accurate and reliable is crucial to successful proteomics research. Complex peptide samples that include a wide range of protein concentrations are a challenge to quantify accurately. According to a recent publication by Hultin-Rosenberg et al. (July 2013), precision and accuracy are important factors that can determine the overall quality of experimental data. Precision is affected by random errors, while accuracy is diminished by systematic errors and the difference in true versus observed values.1
Hultin-Rosenberg et al. investigated a novel method using isobaric tags for relative and absolute quantification (iTRAQ) to precisely and accurately quantify proteins and to help determine which measured protein quantities are accurate. The experimental design was based on duplicate samples in each run that validated each experiment based on its own internal variation of peptide weights, which were in turn based on reporter ion intensity. From those data, peptide weights could be assigned to reflect a high confidence in each protein. The researchers also varied the amount of peptide used and utilized different sample separation methods. Tryptic peptides from A549 lung cancer cells were labeled with iTRAQ and analyzed by three instruments: an LTQ Orbitrap Velos (Thermo Scientific), a 4800 MALDI-TOF/TOF (Applied Biosystems) and a 6350 QTOF (Agilent). This methodology was also demonstrated in a clinical data set of lung cancer tissue samples. The resulting data from each platform were then compared. Hultin-Rosenberg et al. also considered how experimental variables such as sample amount, sample fractionation and fragmentation energy can affect the variance and bias in iTRAQ reporter ions.
The bias and variance in each experiment were mathematically calculated using root mean square error (RMSE), which measures the average magnitude of the error per peptide over all eight iTRAQ channels. Results of the RMSE were plotted against the iTRAQ reporter ion with the lowest intensity. Similar RMSE values were calculated for Orbitrap and MALDI, whereas the QTOF instrument had higher RMSEs.
Comparison of the three instruments determined that the Orbitrap Velos instrument identified approximately four times more proteins. Researchers determined that the optimal mass spectrometry settings on the Orbitrap included a normalized collision energy of 37.5 and a fragmentation time of 30 ms, with the number of target ions set to 50,000. They also found that increasing the amount of loaded peptides four-fold increased the number of identified peptides and proteins in all settings except for the long liquid chromatography gradient (240 min) using the Orbitrap.
The researchers also performed comparisons of the weighted mean (accounts for errors introduced by low-intensity errors), regular mean, and a filtered mean (filtering out low-intensity peptides) for protein quantification. The results of these comparisons revealed that the protein quantities calculated from the weighted mean have smaller relative error than protein quantification calculated from the regular mean. Although the demonstrated improvement was low — <5% — this meant the difference of 90 proteins, which could be a significant loss in experiments to identify prospective biomarkers. Hultin-Rosenberg et al. are confident that the methods and algorithms used in their publication will improve the quality control of protein data and reduce errors during data analysis.
Reference
1. Hultin-Rosenberg, L., et al. (2013, July) “Defining, Comparing, and Improving iTRAQ Quantification in Mass Spectrometry Proteomics Data,” Molecular and Cellular Proteomics, 12(7) (pp. 2021–31).
Post Author: Emily Humphreys. As an undergraduate studying biology at the University of Utah, Emily balanced a heavy class schedule while working long hours in a lab studying eye development. Following graduation, she became involved in infectious disease and aging research involving SNPS.
While she enjoyed the thrill of research, Emily has since traded bench work for science journalism.
She has spent the last year writing about new developments involved in proteomics research, and now food testing.
When she isn’t writing,she can be found playing outside with her kids.
Leave a Reply