PEPPeR, a Platform for Experimental Proteomic Pattern Recognition

Abstract

Quantitative proteomics holds considerable promise for elucidation of basic biology and for clinical biomarker discovery. However, it has been difficult to fulfill this promise due to over-reliance on identification-based quantitative methods and problems associated with chromatographic separation reproducibility. Here we describe new algorithms termed "Landmark Matching" and "Peak Matching" that greatly reduce these problems. Landmark Matching performs time base-independent propagation of peptide identities onto accurate mass LC-MS features in a way that leverages historical data derived from disparate data acquisition strategies. Peak Matching builds upon Landmark Matching by recognizing identical molecular species across multiple LC-MS experiments in an identity-independent fashion by clustering. We have bundled these algorithms together with other algorithms, data acquisition strategies, and experimental designs to create a Platform for Experimental Proteomic Pattern Recognition (PEPPeR). These developments enable use of established statistical tools previously limited to microarray analysis for treatment of proteomics data. We demonstrate that the proposed platform can be calibrated across 2.5 orders of magnitude and can perform robust quantification of ratios in both simple and complex mixtures with good precision and error characteristics across multiple sample preparations. We also demonstrate de novo marker discovery based on statistical significance of unidentified accurate mass components that changed between two mixtures. These markers were subsequently identified by accurate mass-driven MS/MS acquisition and demonstrated to be contaminant proteins associated with known proteins whose concentrations were designed to change between the two mixtures. These results have provided a real world validation of the platform for marker discovery. There is tremendous interest in the use of mass spectrometry as a quantitative technology to measure peptide and protein abundances for comprehensive, system-wide biological research (1, 2). Quantitative proteomics may be used to systematically identify and quantify proteins and their modifications as a function of cell cycle, differentiation, or chemical treatment to obtain novel insights into basic cellular biology. Proteomics also holds promise for discovery of proteins in readily accessible biofluids that are diagnostic or prognostic of a disease condition. Such proteins are termed "biomarkers."The need for robust methods to obtain relative quantification is particularly acute in proteomics-based biomarker discovery where comparative data across multiple patient samples should be obtained (3). Biomarker discovery usually uses biofluids that greatly increase the magnitude of the challenge for quantitative proteomics due to the very high dynamic range of protein abundance (Ϸ10 12 for blood) and the enormous diversity of proteins present in such samples (4).Currently available MS platforms for quantitative proteomics fall roughly into three categories: 1) identity-ba...

A Platform for Accurate Mass and Time Analyses of Mass Spectrometry Data

“…The workflow described here contains functionality similar to traditional AMT 5 and also PEPPeR, 8 but msInspect/AMT was designed to provide a highly flexible approach to combining data from different experimental and biological sources. For example, our AMT matching procedures described above were designed to function when only a potentially small number of peptide locations from an LC-MS experiment are present in the AMT database, such as may be the case when interrogating across dissimilar biospecimen types.…”

“…7 More recently, novel commercial instrumentation such as the LTQ-FT and Orbitrap, and the Q-TOF, have allowed simultaneous acquisition of both high-resolution LC-MS and LC-MS/MS data in a single experiment. Jaffe et al recently described a workflow incorporated in PEPPeR (Platform for Experimental Proteomic Pattern Recognition) 8 that combines peptide locations with amino acid sequences acquired from a series of related runs with these platforms. This workflow is related to the classic AMT approach in that it combines LC-MS and LC-MS/MS data, but because it takes advantage of embedded MS/MS data, it is useful for combining data across a series of related experiments and does not involve the use of externally derived data sources.…”

A strategy for the prior processing of high‐resolution mass spectral data obtained from high‐dimensional combined fractional diagonal chromatography

“…To quantify the sensitivity of the COFRADIC methodology, i.e., the ability to detect low-abundant as well as high-abundant peptide peaks, we used the dynamic range expressed in decibels (dB), which was calculated as: 10 log 10 P max P min (8) with P max and P min denoting the maximum and minimum peptide abundance measured in an experiment, obtained via Eqn (5). A dynamic range of 37 dB was found for all COFRADIC replicates.…”

“…First, available software packages, such as, e.g., SpecArray, [3] PEPPeR, [5] and SuperHirn, [6] which can be used to analyze, combine, and interpret data from high-dimensional LC and MS as an automated procedure, are based on datadriven methods or image processing methods to extract the peptide features from LC-MS. We present an approach that does not operate on the LC-MS image and we argue that the prior processing of LC-MS data can be split into two parts. In the first part, the MS-scans are processed separately, in order to extract peptide features on the basis of prior biological knowledge about the peptide's isotopic distribution, and not on a data-driven method.…”

Quantitative strategies to fuel the merger of discovery and hypothesis-driven shotgun proteomics

“…In general, these computational methods can be divided into two groups. In the first group, spectral features (typically isotopic clusters indicating a peptide) are detected first and then statistically compared [25][26][27][28][29]. Alternatively, the second group applies statistics first so that only differential features are detected from the extracted ion chromatograms [30][31][32][33].…”