Methods In Mass Spectrometry

点击翻译按钮获取中文摘要

Mass spectrometric methods for the analysis of protein structures are being improved, applied and tested in independent and collaborative research projects in proteomics. Projects in separation methods, derivatization chemistry, and data processing are in progress. Peptide analyses by LC-MS/MS generate large datasets, imposing labor-intensive efforts to consolidate peptide and/or protein identification information into meaningful knowledge. Software bioinformatics tools can facilitate and accelerate high-throughput peptide data analyses. Toward this goal, we are developing integrated workflow software designed to maximize the amount of peptide sequence information from LC-MS/MS data. This strategy integrates several independent data search and data mining technologies using a relational database architecture. An iterative search tool acquires results from database search algorithms; spectral filtering and de novo sequencing algorithms are integrated to maximize spectral assignment. These open source software tools will provide an automated data processing/analysis environment for researchers. Probability-based search algorithms calculate a score to indicate the statistical significance of a match between an observed spectrum and that calculated from a sequence search library. Multiple search engines may increase peptide sequence search coverage and/or identify conflicting or ambiguous spectral assignments. We added new features to an open source software package DBParser to combine and validate the search results from OMSSA, Mascot, and X!Tandem. DBParser produces combined or compared peptide or protein lists for subsequent evaluation by research scientists. Protein summaries are based upon the principle of parsimony, identifying the minimal set of proteins that can explain all of the observed peptides. The program generates html output reports of the sorted and compared peptide/protein lists which can be used for subsequent analysis. A CGI-based graphical user interface allows execution of DBParser over a network using a web browser to facilitate its use by multiple laboratories. DBParser has been refined to include quantification. A new technique to determine protein stoichiometry in macro-protein complexes has been tested and is being refined. Protein complexes isolated from mitochondria or sub-cellular organelles are being explored to determine the best methods of isolation and analysis. A new fluorescence detector designed and constructed at NIH is being tested with fluorescent peptide derivatives for quantification and comparison of complex mixtures.