The Clinical Proteomic Technologies for Cancer Initiative, 2006 - 2011, was developed to address the pre-analytical and analytical variability issues that are major barriers to the field of proteomics. These barriers were: (1) experimental design; (2) technological and technical aspects of protein identification; (3) variability related to biospecimens collection; (4) the processes of data acquisition, analysis, and reporting; (5) the lack of reproducible proteomic technologies; and (6) the lack of highly characterized and standardized reagents.
The Unbiased Discovery Working Group attempted to identify the components of the NCI-20 test sample, a mixture containing 20 human proteins, using different mass spectrometry experimental platforms. Three samples of intact NCI-20 (1A) and three samples of trypsin digested NCI-20 (1B) were sent to each laboratory during each of two successive weeks to identify the proteins using their own protocols with any available instruments. Instrument platform diversity was highest in this initial study.
Click to access data
Ref: Tabb DL et al., (2010) J Proteome Res 9(2):761-776
Ref: Wang et al., (2014) Anal. Chem. DOI: 10.1021/ac4034455
The Unbiased Discovery Working Group attempted its first use of a Standard Operating Procedure (SOP v1.0) in re-analyzing the trypsin-digested NCI-20 mixture at sites that had installed Thermo LTQ or Orbitrap instruments. Three “1B” samples were provided to each participating laboratory on two successive weeks. The use of an SOP (v1.0) controlled the rate of MS/MS acquisition but proved less effective in generating similar identified peptide counts across instruments.
Click to access data
Ref: Tabb DL et al., (2010) J Proteome Res 9(2):761-776
Significant changes from Study 3 include; a new Saccharomyces cerevisiae (yeast) reference proteome, the SOP version 2.0, and a bioinformatic infrastructure to collect raw data files and to identify peptides and proteins. Study 3 tested these tools in a small-scale methodology test. The Unbiased Discovery Working Group tested a new version of the SOP v2.0 for the complex yeast lysate sample introduced in this study. Each instrument was evaluated for SOP v2.0, specifying 184-minute RPLC analyses, in two replicates of NCI-20 followed by yeast lysate. For this study only, an Applied Biosystems QSTAR-Elite accompanied the LTQ and Orbitrap instruments.
Click to access data
Ref: Tabb DL et al., (2010) J Proteome Res 9(2):761-776
Ref: Paulovich AG et al., (2010) Mol Cell Proteomics 9:242-254
Performed by the CPTAC Experimental Design and Statistics: Verification Studies Working Group assessed the utility of multiple reaction monitoring (MRM) coupled with stable-isotope dilution mass spectrometry (SID-MS) for quantification of proteins in human plasma with a particular focus on the reproducibility and transferability of protein-based MRM assays across multiple laboratories. Towards this goal, CPTAC Study 4 was implemented to assess intra- and inter-laboratory performance of an MRM/SID-MS assay designed to quantitate 10 target peptides representing 7 target proteins that were spiked into human plasma. This initial (unpublished) “feasibility” study developed standard operating procedures (SOPs) for sample preparation, data acquisition and analysis. A follow-up study that systematically evaluated the effect of sources of variability within the study design, including spiking of the 7 target proteins into human plasma and digestion at each individual site is described in Study 7 (see below).
Study 5 extended upon Study 3 by revising the SOP to version 2.1 and adding a new sample: yeast sample spiked with BSA. The Unbiased Discovery Working Group probed the yeast sample in depth and evaluated the impact of spiking a small amount of BSA into samples. Each LTQ and Orbitrap instrument produced six RPLC analyses of yeast and six of BSA-spiked yeast, with NCI-20 samples present as QC mixtures. The study showed no negative effects from the spikes on other identifications, but the need for SOP v2.1 to specify flow rate was demonstrated by an outlier instrument.
Click to access data
Ref: Tabb DL et al., (2010) J Proteome Res 9(2):761-76
Ref: Rudnick PA et al., (2010) Mol Cell Proteomics 9:225-241
Ref: Wang et al., (2014) Anal. Chem. DOI: 10.1021/ac4034455
Study 6 built upon Study 5 by including spikes of the Sigma UPS 1, a mixture of 48 human proteins in equimolar concentration. The Unbiased Discovery Working Group evaluated the sensitivity of spiked protein detection by evaluating samples of yeast with Sigma UPS 1 spiked at five levels. The yeast, Sigma UPS 1, and five spiked levels were each analyzed three times by RPLC, and NCI-20 samples were present as QC mixtures. The Sigma UPS 1 proteins were not detected at the lowest concentration, but each of the other concentrations generated UPS 1 detections, increasing with concentration. This Study reflects the final SOP v2.2 specification for LTQ and Orbitrap instruments.
Click to access data
Ref: Tabb DL et al., (2010) J Proteome Res 9(2):761-76
Ref: Rudnick PA et al., (2010) Mol Cell Proteomics 9:225-241
Ref: Paulovich AG et al., (2010) Mol Cell Proteomics 9:242-254
Executed by the CPTAC Experimental Design and Statistics: Verification Studies Working Group, was a multi-laboratory study designed to assess performance metrics of multiplexed, protein-based Multiple Reaction Monitoring (MRM) assays, including recovery, precision, and limits of detection and quantitation. Although individual laboratories have demonstrated that MRM coupled with stable-isotope dilution mass spectrometry has suitable assay performance for quantitative measurements of candidate protein biomarkers in plasma, reproducibility and transferability of these assays across multiple laboratories has yet to be demonstrated. CPTAC Study 7 was an advanced consortium-wide (8 laboratories participated) follow-up study to the initial “feasibility” Study 4. Study 7 encompassed a three-tiered experimental protocol that progressively introduced sample preparation variables likely to affect inter- and intra-laboratory reproducibility, transferability, and sensitivity. The experimental design varied from 1) spiking digested plasma with 10 signature peptides representing 7 target proteins (samples prepared centrally at NIST), 2) spiking digested plasma with the 7 digested target proteins (samples prepared centrally at NIST), and finally 3) spiking the 7 intact target proteins into un-digested, neat plasma followed by combined digestion (samples prepared at each participating site). In Study 7, using common materials and standardized protocols, it was demonstrated that MS-based assays of proteins in plasma can be sensitive and highly reproducible across laboratories and instrument platforms. Here we provide the Study 7 data files to be used as benchmarks against which individual laboratories can compare their performance and evaluate new technologies for biomarker verification in plasma.
Click to access data
Ref: Addona TA et al., (2009) Nat Biotechnol 27:633-641
The Unbiased Discovery Working Group re-evaluated the yeast lysate sample without the requirement of an SOP. Each group was asked to produce three RPLC separations of the yeast sample at both high and low concentrations using their own protocols. The same quantity of yeast was analyzed in Studies 5, 6 and 8, with an additional 5x (high load) sample analyzed in Study 8. The LTQ and Orbitrap instruments generated a broad range of identified peptides as groups used their own lab-specific protocols.
Click to access data
Ref: Tabb DL et al., (2010) J Proteome Res 9(2):761-776
Ref: Paulovich AG et al., (2010) Mol Cell Proteomics 9:242-254
The Post-Translational Modification Working Group evaluated the reproducibility of a lectin-based glycopeptide enrichment and identification workflow on the conditioned medium of luminal and triple-negative breast cancer cell-lines. Three sites and two different instrument platforms analyzed lectin-enriched and PNGase-F treated aliquots from digested conditioned media (CM) by LC-MS/MS.
The CM samples were initially trypsin-digested and aliquoted at a single site and then distributed to 3 sites. Each site separated the 10 CM samples, in duplicate, by chromatography using lectins Aleuria aurantia (AAL) and Sambucus nigra agglutinin (SNA), which generated 40 fractions. The samples were deglycosylated using PNGaseF and analysed in duplicate, yielding a total of 80 MS/MS data sets per site. The lectin affinity chromatographic enrichment and LC-MS/MS protocol can be accessed from here.
N-deglycosylated glycopeptide identifications provided on N-glycosylation sites and N-Glycosites were compared between sites and breast-cancer derived samples.
Click to access data
Ref: Drake et al., (2012) J Proteome Res 11 (4): 2508–2520