Session TOC. There are 4 abstracts in this session.



Session: COMPUTATION & ANALYSIS: COMPUTATIONAL AND STATISTICAL METHODS, time: 09:50 AM - 10:15 AM

Intelligent Data Acquisition for Multiplexed Proteomics

Devin Schweppe

High-throughput quantitative proteomics requires efficient sample handling, fast mass spectrometric acquisition, and precise quantitation. Modern methods for multiplexed proteomics enable the utilization of isobaric chemical labeling to compare up to 16 samples simultaneously. However, acquisition of multiplexed proteomic quantitative data generally suffers from either slow acquisition speeds (SPS-MS3) or limited dynamic range (HRMS2). To bridge this divide we developed intelligent data acquisition strategies to acquire high accuracy quantifying scans based on real-time spectral matching. Real-time database searching of spectra during data acquisition enabled a 100% increase in data acquisition speed and improved quantitative accuracy. When applied to diverse human cell lines models, the real-time search strategy reproducily quantified 8000-9000 proteins enabling robust comparison of complex samples. Intelligent data acquisition in combination with quantitative proteomics offers new avenues to explore and understand the proteome across conditions, tissues of origin or species of interest.
Tips and Tricks:

Session: COMPUTATION & ANALYSIS: COMPUTATIONAL AND STATISTICAL METHODS, time: 10:15 AM - 10:40 AM

Measuring phosphorylation state in the human proteome

Brian Searle

While mass spectrometry has made a far-reaching impact towards understanding cellular signaling, there is still a huge limitation in analyzing phosphosites occurring in close proximity. Indeed, accumulating phosphoproteomic data shows that phosphorylation sites cluster together in multi-phosphorylated proteins, where over half of sites are within four amino acids of each other. These neighboring sites result in phosphopeptide positional isomers that can sometimes be chromatographically resolved, but because they have the same precursor mass, dynamic exclusion settings often cause these peptides to be overlooked in data-dependent acquisition (DDA) experiments. This, coupled with the stochastic nature of DDA, often results in replicate quantitative experiments that exhibit very poor overlap. Here we propose Thesaurus, a new search engine that detects clusters of phosphopeptide positional isomers from Parallel Reaction Monitoring (PRM) and Data-Independent Acquisition (DIA) experiments. Using the insulin signaling pathway as a model, we demonstrate we can computationally extract distinct quantitative signaling effects of different positional isomers, even if those isomers do not separate chromatographically.
Tips and Tricks:

Session: COMPUTATION & ANALYSIS: COMPUTATIONAL AND STATISTICAL METHODS, time: 10:40 AM - 10:55 AM

Use of Gas Phase Fractionation Data Independent Acquisition and Spectrum Centric Searching to Build Matrix Specific Libraries for Quantitative Analysis

Lilian Heil; Michael Maccoss
University of Washington, Seattle,

Due to significant advances in mass-spectrometry instrumentation, data-independent acquisition (DIA) has emerged as a powerful acquisition scheme to sample all peptides in a specified mass range. Because of limitations in scan speed, there is a trade-off between the mass range covered and the width of the isolation window. Wider isolation windows offer more complete mass-range coverage but highly chimeric spectra may complicate data analysis. To address the challenges of complex spectra, these data are normally analyzed using a peptide-centric strategy, making use of prior information about 1) which peptides are likely to be found in the sample, 2) the expected retention time, and 3) fragment ion intensity. Bruderer et al. showed that it was valuable to produce matrix specific libraries to perform this analysis to limit searches to peptides likely present in the sample and to improve the statistical power. A limitation of this strategy is that library information is collected by biochemically fractionating the sample and collecting spectra using data dependent acquisition (DDA). We and others have shown that DDA is a poor predictor of the best peptides to use for quantitative analysis. Additionally, the use of a fractionated sample means that indexed retention time prediction will be performed in a different matrix than the unfractionated samples used for quantitation. Searle et al. expanded upon this concept by using gas-phase fractionation to acquire a small number of narrow window DIA runs of a pooled sample to facilitate sample-specific library preparation. Library assembly from these narrow window data typically implements peptide-centric approaches or builds off existing spectral libraries, but these searches have limited sensitivity, especially in detecting post-translational modifications.  Here, we implement a sensitive spectrum-centric search approach to generate sample-specific spectral libraries and demonstrate the ability of these libraries to increase peptide detections and improve quantitative analyses in biological samples.

Tips and Tricks:

Session: COMPUTATION & ANALYSIS: COMPUTATIONAL AND STATISTICAL METHODS, time: 10:55 AM - 11:10 AM

MassIVE.quant: a community resource of curated quantitative mass spectrometry-based proteomics datasets

Meena Choi1; Jeremy Carver2; Cristina Chiva3; Manuel Tzouros4; Ting Huang1; Tsung-Heng Tsai1; Benjamin Pullman2; Oliver M. Bernhardt5; Ruth Hüttenhain6; Guo Ci Teo7; Maria Pavlou8; Erik Verschueren6; Bernd Wollscheid8; Alexey Nesvizhskii7; Lukas Reiter5; Tom Dunkley4; Eduard Sabidó3; Nuno Bandeira2; Olga Vitek1
1Northeastern University, Boston, MA; 2University of California, San Diego, San Diego, CA; 3Center of Genomics Regulation, Barcelona, Spain; 4Hoffmann-La Roche Ltd, Basel, Switzerland; 5Biognosys, Zürich, Switzerland; 6University of California, San Francisco, San Francisco, CA; 7University of Michigan, Ann Arbor, MI; 8Institute of Molecular Systems Biology, ETH, Zürich, Switzerland

We present MassIVE.quant (http://massive.ucsd.edu/ProteoSAFe/static/massive-quant.jsp), a tool-independent repository infrastructure and data resource for reproducible quantitative mass spectrometry-based biomedical research. MassIVE.quant is an extension of MassIVE (the Mass Spectrometry Interactive Virtual Environment) to provide the opportunity of large-scale deposition of heterogeneous experimental datasets and facilitate a community-wide conversation about the necessary extent of experiment documentation and the benefits of its use. It supports various reproducibility scopes, such as the infrastructure to fully automated the workflow, to store, and to browse the intermediate results. First, MassIVE.quant supplements the raw experimental data with detailed annotations of the experimental design, analysis scripts, and results, which enable the quantitative interpretation of mass spectrometry-based experiments and the online interactive exploration of the results. A branch structure enables to view and even compare reanalyses of each experiment with various combinations of methods and tools. Second, the curated alternative workflows can be used off-line and online reanalyses of the data starting from an intermediate output in MassIVE.quant. MassIVE.quant is independent of data acquisition types and of computational tools used to complete the analyses. To exemplify the utility of storing, sharing, reanalyzing, and curating data from quantitative experiments, we present the first compilation of proteomic datasets from benchmark controlled mixtures and biological investigations, interpreted with various data processing tools and analysis options. The extensive documentation for workflow and data submission, including video tutorials, is available.

Tips and Tricks: