Factors essential for optimization of automated electron transfer dissociation (ETD)-based mass spectral peptide identification: Implementation of Charge-State and Sequence-Dependent Scoring.

Alma Burlingame
Professor
Chemistry and Pharmaceutical Chemistry
University of California, San Francisco, USA

Abstract: The recent advent of electron transfer dissociation (ETD) fragmentation is providing a new powerful and versatile dimension for orthogonal structural characterization of peptides and small proteins observed in the eluant of liquid chromatography tandem mass spectrometry experiments. Since the early 1990’s peptide sequence identification has been performed with the aid of search engines that attempt to match measured CID spectra of peptides to putative partial protein sequences in a genomic and protein databases. However, our knowledge of the salient factors necessary to match ETD fragmentation with these databases is currently less well developed than common algorithms used for the analysis of the more ubiquitous collision induced dissociation(CID) fragmentation spectra.
In this study we report a a new scoring system(goodness of experimental match vs database entries) developed for analysis of peptide ETD fragmentation spectra that incorporates information on ion type weighting depending on the precursor ion charge state and peptide sequence. Evaluation of the performance of this new scoring regime was carried out using both previously published results from alternative search engines and datasets from our own laboratory. Using this new scoring we report that Protein Prospector is able to identify 60% more spectra than the other most successful program at the same defined false discovery rate threshold. This result suggests that other search engines would benefit from adoption of similar scoring considerations for use with ETD and ECD peptide spectra.

Short Bio: Dr. Burlingame is a Professor of Chemistry and Pharmaceutical Chemistry at University of California, San Francisco and the Director of the NCRR National Research Resource in Mass Spectrometry and Proteomics at UCSF. The Center's purpose is to promote the effective usage of leading techniques of mass spectrometry in biomacromolecular research and to solve the challenging proteomics problems that impinge on molecular abnormalities that underlie human disease. His main interest is in using mass spectrometry to study epigenetic modulation and regulation of mammalian and human proteomes, with the hope of better understanding the cause of diseases and cancer.

Instrument and methods development for characterization of protein and peptide post-translational modifications by mass spectrometry.

Donald F. Hunt
Professor
Biochemistry & Molecular Biology
University of Virginia, USA

Abstract: This lecture will focus on chemical enrichment technology, peptide derivatization, and mass spectrometry instrument development that facilitates identification of posttranslational modifications on proteins and peptides. Development of a new ion source that facilitates simultaneous generation of positive and negative ions by electrospray ionization and chemical ionization, respectively, now makes it possible to record collision activated dissociation (CAD) and electron transfer dissociation (ETD) mass spectra on the high resolution LTQ-FT instrument. Parent ion and fragment ion mass spectra can be obtained at resolutions in excess of 40,000 and with mass measurement accuracy in the high ppb range. Applications of this instrumentation for the characterization of class I MHC phosphopeptides, and O-GlcNacylated proteins will be described.

Short Bio: Prof. Hunt’s 40-year career as a pioneer in the field of mass spectrometry saw him publish more than 300 articles and send more than 100 former graduate and postdoctoral students to leading positions in the field of mass spectrometry. Prof. Hunt is widely recognized for outstanding contributions to the field of mass spectrometry, especially in the micro sequencing of proteins, electron transfer dissociation (ETD) and electron-capture negative-ion mass spectrometry. Widely regarded in the proteomic community, Prof. Hunt has been awarded several honors including the Distinguished Contribution Award from the American Society of Mass Spectrometry, the Christian B. Anfinsen Award from the Protein Society, the Chemical Instrumentation Award and Field and Franklin Award from the American Chemical Society, the Thomson Medal from the International Mass Spectrometry Society, the Human Proteome Organization Distinguished Achievement Award in Proteomics, and most recently the Association of Biomolecular Resource Facilities 2007 Award.

Scoring Identification and Localization of Phosphorylation Sites in Peptide MS/MS Spectra.

Karl Clauser, Ph.D.
Research Scientist
Proteomics
The Broad Institute of MIT and Harvard

Abstract: Scoring the matches of an MS/MS spectrum to a peptide sequence in the most widely used search engines typically deals primarily with identifying the right peptide. A few researchers active in phosphoproteomic work have built separate auxiliary programs that focus additional scoring on assessing the certainty or ambiguity in localizing the site(s) of phosphorylation to particular Ser, Thr, or Tyr residues in the peptide. With a focus on phosphorylation, this talk will describe the development of an algorithm within the Spectrum Mill search engine for scoring the localization of variable modifications. Performance of the scoring algorithm will be presented in the context of the ABRF iPRG 2010 study where 33 participants all independently analyzed a common phosphoproteomic dataset generated from the tryptic digest of a lysate of human K562 cells following SCX/IMAC phosphopeptide enrichment. The talk will further seek to stimulate discussion on the issue of how we can standardize the calculation of a false localization rate for automated phosphosite assignment not only for comparing performance of different scoring algorithms but also for stating the quality of data analysis when publishing phosphoproteomic research.

Short Bio: Karl Clauser is an internationally recognized leader in proteomics software development and phosphoproteomics. He has applied advanced separations methods, mass spectrometry and informatics to characterize protein pathways, understand the mechanism of action of drug candidates, define biomarkers of disease and drug effect, and build understanding of protein targets and their roles in disease in proteomics groups at Genentech and Millennium Pharmaceuticals prior to joining the Broad. As part of his doctoral studies he co-authored Protein Prospector, a widely used proteomics software package accessible freely over the Web. He also the developer of SpectrumMill, acknowledged as one of the three most powerful software packages for proteomics data analysis.

Martin McIntosh
Principal Investigator
Computational Proteomics Laboratory
Fred Hutchinson Cancer Research Center (FHCRC)

Abstract: Mass spectrometry with stable isotope labeled biomaterials is possibly the most common approach to compare two experimental conditions. It has been under appreciated, however, that in these experiments even a small numbers of peptide sequence errors or other signal processing errors can cause bias. It may not be uncommon for a large fraction of all up regulated proteins identified in an experiment to be incorrect, and the error rate among those proteins will exceed the peptide-level error rate used in the search protocol. Many of these errors can be easily identified by visual inspection even by individuals with only modest training. Large-scale visual inspection is ordinarily impractical, but we will present and possibly demonstrate a tool that allows people to inspect relevant quantitative events, eliminate bad ones and then automatically recalculate the results. We also summarize our exploration of using crowd sourcing to curate large-scale experiments.

Bio: Martin McIntosh, Ph.D., is a Full Member at the Fred Hutchinson Cancer Research Center (FHCRC) and Principal Investigator of the Computational Proteomics Laboratory (CPL). Dr. McIntosh has a long history of research in biomarkers and early detection of disease, especially of ovarian cancer. He is the leader of an early detection project of the FHCRC's ovarian cancer SPORE award and is one of three PIs of the NCI's proteomics initiative consortia. Dr. McIntosh is PI of the Early Detection Research Network (EDRN) biomarker development laboratory (for breast and ovarian cancer), and his laboratory leads the data integration and mining for several consortia for cancer as well as neurodegenerative disease research. His primary research focus involves discovery and evaluation of biomarkers and biomarker panels for early disease detection. He is an Associate Editor for the Journal of Proteome Research and co-Chair of the biomarker initiative for the Human Proteome Organization (HUPO). Dr. McIntosh serves in leadership positions both at the FHCRC and nationally for biomarker discovery research programs.

Computational tools and crowd sourcing to improve quantitative proteomics experiments.

Michael MacCoss
Associate Professor
Genome Sciences
University of Washington, USA

Abstract: Mass spectrometry with stable isotope labeled biomaterials is possibly the most common approach to compare two experimental conditions. It has been under appreciated, however, that in these experiments even a small numbers of peptide sequence errors or other signal processing errors can cause bias. It may not be uncommon for a large fraction of all up regulated proteins identified in an experiment to be incorrect, and the error rate among those proteins will exceed the peptide-level error rate used in the search protocol. Many of these errors can be easily identified by visual inspection even by individuals with only modest training. Large-scale visual inspection is ordinarily impractical, but we will present and possibly demonstrate a tool that allows people to inspect relevant quantitative events, eliminate bad ones and then automatically recalculate the results. We also summarize our exploration of using crowd sourcing to curate large-scale experiments.

Short Bio: The focus of The MacCoss Group's research is in the development of stable isotope and mass spectrometry based approaches to improve the understanding of biology on a molecular, cellular, and whole organism level. Presently, individuals in the laboratory are working on technology for 1) automating biochemical sample preparation methods for the analysis of protein mixtures; 2) developing in vivo stable isotope methods for studying protein metabolism; 3) increasing the dynamic range of liquid chromatography-mass spectrometry for the analysis of peptides; and 4) developing computational tools for the automated conversion of mass spectrometry data into biologically meaningful results. These technologies are presently being demonstrated in the model organisms C. elegans and S. cerevisiae. Although their current research interests are presently in model systems, their long-term goal is have technologies robust enough to handle the automated high-throughput characterization of human clinical samples.

Neil Kelleher
Professor
Department of Chemistry
University of Illinois at Urbana-Champaign

Abstract: Top Down mass spectrometry (MS) is emerging as an alternative to common Bottom Up strategies for protein analysis. For either strategy, measuring fragment ions with 1-10 ppm mass accuracy results in "Precision Proteomics" (i.e., the hyper-confident assignment of identity and modifications. In the Top Down approach, intact proteins are fragmented directly in the mass spectrometer to achieve both protein identification and characterization, even capturing information on combinatorial post-translational modifications. Just in the past two years, Top Down MS has seen incremental advances in instrumentation and dedicated software, but has also experienced a major boost from refined separations of whole proteins in complex mixtures that have both high recovery and reproducibility. Combined with steadily advancing data processing pipelines, a high-throughput workflow covering intact proteins and polypeptides up to 80 kDa is directly visible in the near future.

Bio: Professor Kelleher received a B.S. and B.A. from Pacific Lutheran U. in 1992, a Fulbright Fellowship the following year, and a Ph.D. from Cornell U. in 1997. After a NIH Postdoctoral Fellowship at Harvard Medical School with Chris Walsh, Kelleher joined the faculty at UIUC in 1999 as a bioanalytical chemist. He has received numerous awards, including the 2009 Biemann medal award for his pioneer role in the development of top-down mass spectrometry instrumentation, methods and bioinformatics.

Interpreting the protein language: Quantitative proteomics approaches for characterization of post-translational modifications.

Ole N Jensen

Professor
Biochemistry & Molecular Biology
University of Southern Denmark, Denmark

Abstract: Proteins are the main workhorses of cells and the spatio-temporal regulation of protein function is the foundation for the dynamic, transient biochemical processes that govern differentiation, proliferation, growth of cells. Post-translational modifications (PTMs) of proteins provides a mechanism for tuning protein activity as a function of the surrounding environment. Mass spectrometry based techniques for mapping and quantitation of PTMs provides an accurate and sensitive approach to study dynamic cellular events, for example in the context of signalling and epigenetic regulation. I will provide an overview of current analytical and computational strategies that are used to study dynamic PTMs using functional proteomics.

Short Bio: Prof. Ole N. Jensen has worked in the area of protein mass spectrometry since 1988 and in proteomics since 1994. He pioneered the application of mass spectrometry to investigations of UV cross-linked protein-nucleic acid complexes (ON Jensen et al, 1993; SE Bennett et al 1994; ON Jensen et al 1996; H Steen et al 2001; H Steen and ON Jensen, 2002) and he is currently extending these studies to address a range of questions in the area of chromatin structure and gene regulation via protein analysis by high-performance tandem mass spectrometry. Ole N. Jensen recently received a young investigator award (2005-2007) from the Danish Natural Sciences Research Council and he was appointed Lundbeck Foundation Research Professor (2005-2009).

LC-IMS-TOF MS Platform for Proteomics; Challenges for Data Processing and Analysis.

Richard Smith
Professor, Chief Scientist (PNNL), Director of the NIH Research Resource for Integrative Proteomics
Chemistry
Pacific Northwest National Laboratory, USA

Abstract: Mass spectrometry based proteomics technologies are powerful discovery-based tools based upon their ability to measure up to thousands of proteins from complex biological samples in a single analysis. In spite of significant advances, MS technologies remain challenged by biofluid samples such as blood plasma where protein concentrations span a dynamic range greater than 10 orders of magnitude and inadequate measurement throughput hinders the ability to effectively account for biological diversity. Achieving higher analysis throughput with concurrent increase in sensitivity and dynamic range of measurements to allow detection and quantitation of lower abundance peptides and proteins from biological fluids is a major analytical challenge. To address this challenge we have been developing a platform that encompasses fast capillary LC separations coupled via a greatly improved electrospray ionization interface to an ion mobility separation (IMS) stage that is interfaced to a time-of-flight mass spectrometer (TOF MS). This presentation will describe the platform, the differences in performance and data qualities compared to other presently popular mass spectrometry platforms, and discuss challenges associated with data processing and analysis.

Short Bio: Dr. Smith's research has involved the development and application of advanced analytical methods and instrumentation, with particular emphasis on high resolution separations and mass spectrometry, and their applications in biological and biomedical research. Much of his research over the last 15 years has focused on the development and application of new ultra-sensitive and comprehensive methods for quantitatively probing the entire array of proteins expressed by a cell, tissue or organism, i.e., their "proteomes". Current interests also include greatly increasing the throughput and sensitivity of proteomics and metabolomics measurements to meet the needs of systems biology research.
Dr. Smith is an adjunct faculty member of the Department of Chemistry, Washington State University, and the Department of Chemistry, University of Utah, and an affiliate faculty member of the Department of Chemistry, University of Idaho. He has presented more than 350 invited or plenary lectures at national and international scientific meetings, and is the author or co-author of more than 575 publications. Dr. Smith holds twenty seven patents and has been the recipient of seven R&D 100 Awards. He is Director of the NIH Research Resource for Integrative Proteomics located at PNNL.

Bioinformatics in mass spectrometry based proteomics: the importance of respecting chemistry and biology.

Roman A. Zubarev
Professor
Division of Molecular Biometry
Institute for Medical Biochemistry and Biophysics, Karolinska Institute, Sweeden

Abstract: The proteomics field driven by the progress in mass spectrometry has greatly benefited from the efforts of bioinformatics experts whose mission is to provide rigorous theoretical basis for data interpretation. However, there is still a gap between the mathematical approaches bioinformatics pursues in solving proteomics problems and the reality that is based on physico-chemical phenomena. Ignoring chemistry and biology can be counter-productive, and even perilous, for beautiful mathematical models built on simplified assumptions. Reality is often more complex than we would like to believe. As a physicists turned chemist turned biologist, the author has a first-hand experience of learning the hard way to respect the wisdom coming from other disciplines, even in the cases when this wisdom needs to be challenged. Several examples will be provided where the conclusions drawn from bioinformatics findings caused controversial reaction among chemists and biologists. In one example, where the fragmentation pathways of tryptic peptide dications were examined, the statistical evidence pointed towards a specific chemical structure of b2 ions, which stirred controversy in the theoretical chemistry community. In another example, some biologists became disturbed by the estimates of the extent of posttranslational modifications in human protein samples. The reversed situations are also known. The validity of the "one peptide - one chromatographic peak" paradigm has only recently become questioned by bioinformatics, while peptide chemists knew this for decades. But with all due respect to chemistry and biology, the bioinformatics insights coming from processing of massive amount of data are often impossible to gain via traditional methods, as exemplified by the paradox hidden in molecular masses of biomolecules consisting of the elements C, H, N and O.

Short Bio: Received Master degree in Technical Physics (specialization – microelectronics) at the Moscow Engineering Physics Institute in 1986. PhD in Ion Physics in 1997 from the Uppsala University, Sweden. The same year moved to Cornell University, USA as a postdoc in Baker Chemistry Lab with Prof. F.W. McLafferty. In 1998 became Associate professor of biological mass spectrometry at the Department of Chemistry of the University of Southern Denmark in Odense. Moved to the Uppsala University in 2002 as professor of Ion Physics and Biological mass spectrometry. In 2007 established a group of Molecular Biometry at the Institute of Cell and Molecular Biology of the Uppsala University. Main awards: Curt Brunnée Award of the International Mass Spectrometry Society “for outstanding contribution to the development of MS instrumentation”, Prague, August 2006. Biemann Medal for “a significant achievement in basic or applied mass spectrometry made by an individual early in his or her career” from the American Society of Mass Spectrometry, Indianapolis, June 2007.

Using literature annotations to improve peptide PTM assignments.

Ronald Beavis

Professor
Medical Genetics
University of British Colombia, Canada

Abstract: Correctly assigning post-translational modifications to peptide sequences can be challenging when considering large lists of modifications because of both computational complexity and the potential for false assignments. We have developed a simple method that allows testing for a comprehensive list of modifications that reduces the complexity of the calculation and the scope for false assignments based on known PTM assignments that have been derived from the literature, bioinformatics predictions and proteomics data repositories.

Short Bio: Dr. Beavis is a member of the Biomedical Research Centre at the University of British Columbia, where he also holds the Tier I Canada Research Chair in Experimental Bioinformatics. He obtained his Ph. D. in Physics at the University of Manitoba in Dr. Ken Standing’s Time-of-Flight lab. Dr. Beavis started his professional career as a NATO postdoctoral fellow at the Technical University of Munich. At Rockefeller University, he discovered most of the matrix compounds currently used for MALDI mass spectrometry and developed a new type of ion source that has become the basis of most commercial MALDI instrument designs. He joined Eli Lilly and Co. in 1997 and formed his own bioinformatics company Proteometrics in 1999. He sold Proteometrics to Genomic Solutions in 2003 and remained in private industry as a consultant until 2006.