Pino LK, Searle BC, Yang HY, Hoofnagle AN, Noble WS, MacCoss MJ. Matrix-Matched Calibration Curves for Assessing Analytical Figures of Merit in Quantitative Proteomics. J Proteome Res. 2020 Mar 6;19(3):1147-1153. doi: 10.1021/acs.jproteome.9b00666. Epub 2020 Feb 24. PMID: 32037841; PMCID: PMC7175947.
Mass spectrometry is a powerful tool for quantifying protein abundance in complex samples. Advances in sample preparation and the development of data independent acquisition (DIA) mass spectrometry approaches have increased the number of peptides and proteins measured per sample. However, methods to assess quantitative figures of merit (e.g. lower limit of quantification, LLOQ) are not easily extended to multiplex assays with hundreds or thousands of analytes. Here we present a series of experiments demonstrating how to assess whether a peptide measurement is quantitative by mass spectrometry. In a study of the yeast proteome, only 52% of detected proteins (41% of detected peptides) have a peptide that is above the LLOQ in the reference material. A similar trend was observed in human cerebrospinal fluid, suggesting that this observation is not sample-specific. Our results demonstrate that increasing the number of detected peptides and proteins in an unbiased proteomics experiment does not necessarily result in increased numbers of peptides or proteins that can be measured quantitatively. Furthermore, our method provides an approach to determining useful figures of merit for hundreds to thousands of proteins in the same experiment.
In this project, we propose a framework for discriminating between detected and quantitative peptides in proteomics experiments. We introduce an alternative to stable isotope label calibration curves called "matched matrix calibration curves, and we demonstrate the method using yeast as a model system. We additionally apply the proposed framework for cerebrospinal fluid (CSF) proteomics and formalin-fixed paraffin-embedded (FFPE) renal biopsies. We find that, regardless of the sample type, the framework shows that not all detected peptides are quantitative, and that peptide LOQs range orders of magnitude even within the same protein. This illustrates the need for additional criteria to ensure that detected peptides are also quantitative. We thus demonstrate that proteomics experiments employing matched matrix calibration curves can produce truly quantitative measurements of peptides and proteins at the scale of an entire proteome.