Paul Rudnick National Institute of Standards and Technology (NIST) Office: 301-975-5825 FAX: 301-869-4020 E-mail: paul.rudnick@nist.gov Job Title: Scientist Ph.D. in Microbiology from the University of Arizona

Speaker: Paul Rudnick, Mass Spectrometry Data Center, National Institute of Standards and Technology, 100 Bureau Drive, Gaithersburg, MD 20899

Topic: Reference Libraries of Peptide Tandem Mass Spectra: 2007

Place: Building 549, Auditorium, NCI at Frederick, Frederick, MD

Time: Tuesday, September 18, 2007, at 2:00 PM

Abstract: In June 2006 the first versions of the NIST MS/MS peptide fragmentation libraries were publicly released and the methods for building and annotating consensus spectra were described. The primary motivator for building and using these libraries in proteomics research is that the library spectra are generated directly from experimental data and are thus more likely to resemble test spectra than theoretical spectra predicted from protein sequence. This is increasingly true as greater numbers of replicate spectra are processed and new versions of the libraries are released. Additionally, scoring algorithms can make extensive use of peak intensities, neutral loss peaks and less predictable fragmentation patterns as discriminators, increasing the sensitivity and specificity over sequence searching alone. In this report, we will focus on (1) the growth of the major species libraries over the past year due to the inclusion of spectra from an increasing number of public and private sources, including 2 major HUPO projects. A July 2007 build of the yeast ion trap library added approximately 44,000 new consensus spectra, increasing the total content more than 2-fold. Searches against this library commonly yield more identifications than sequence searches for 1D shotgun experiments. A similar or greater increase is expected for the next human build. Statistics covering the numbers of spectra processed, the number of peptide ions included in each library, and the relative coverage versus theoretical digests of each proteome versus time will be presented. (2) Practical methods for batch searching these libraries alone or in combination with OMSSA (NCBI) will also be presented as will a method for determining false discovery rates using a library from a taxonomically distinct organism. (3) Future expansion of the libraries, methods to improve coverage of PTMs and low abundance peptides, and mechanisms for releasing these libraries will be discussed.

The slides to this seminar are in a 1.1 Megabyte PDF file, which can be opened and read by using the free Adobe Acrobat Reader®.