The University of Texas at Austin is awarded a grant to develop new algorithms and validate a software system for large-scale protein identification and measurement of protein expression level by mass spectra look-up. These algorithms will be integrated into MoBIoS, a metric-distance database management system dedicated to the management of biological data. Mass-spectroscopy analysis of the proteins of a cell lysate often yields >30,000 spectra. Current analysis methods iterate through the spectra. Metric-distance join algorithms on sets of experimental mass spectra promise order-of magnitude performance improvements. Mass spectral data will be expanded and integrated into the capabilities in the Open Proteomic Database (OPD) thereby building a reference library of experimental peptide mass spectra. These spectra can be used to supplement or supplant the use of the approximate, computationally predicted peptide mass spectra currently used for spectral look-ups in proteomics, thereby further improving the ability to identify peptide spectra in proteomics experiments. The ultimate goal of this project is to integrate OPD and MoBIoS to create a MoBIoS-based spectral matching application and enable spectral queries of OPD. This will allow mapping of all spectra in an experiment against all experimental spectra in OPD (i.e., reference-library based analysis of proteomics data). More generally, this will allow discovery of recurring experimental spectra, leading to better rules for computationally generating peptide fragmentation spectra, as well as to identification of putative post-translational modifications and mutations independently of predicted reference spectra. The team will participate in the Computer Science department's First Bytes program, a co-ed summer camp that encourages high school students to major in computer science. The integration of Computational Biology material into this program is part of an explicit effort to broaden the appeal of Computer Science by providing a face to computers that contrasts to computer games. Both platform and application software will be made available for open academic use. The protein quantitation analysis application achieves its result without the use of isotope labeling. Thus, laboratories that cannot afford these proprietary reagents will be able to conduct quantitation experiments.
