Plenary Lectures L2.1 Homology driven proteomics and its implications for animal and plant biology

Homology driven proteomics enables the biochemical characterization of unconventional, yet biologically interesting model organisms from a variety of taxonomic classes, including plants, fungi, insects, reptiles, among many others. Confident identification of unknown (i.e. not present in a database) proteins relies upon often marginal similarity between sequences of fragmented peptides and known homologous sequences from phylogenetically related species. A single dataset of MS/MS spectra is acquired at the uncompromised sensitivity and database searches are performed in a layered manner. Stringent searches (MASCOT), powered by the high mass resolution and accuracy of LTQ Orbitrap machine, rapidly identify most conserved proteins sharing a few identical peptide sequences with known proteins. The remaining unmatched MS/MS spectra are filtered against a library of non-annotated background MS/MS spectra and then rapidly interpreted de novo in a batch mode. Several candidate sequences that might be only partially accurate and highly redundant are produced for each fragmented precursor. Hence, the interpretation of a typical LC-MS/ MS run typically yields several thousands peptide sequence candidates, which are merged into a single search string and submitted to the Mass Spectrometry driven BLAST (MS BLAST) server at http://genetics.bwh.harvard.edu/msblast/index.html. MS BLAST searches enable confident identifications of unknown proteins in crude mixtures, such as insect and reptile venoms, total extracts from plant seeds etc. In organisms with known genomes, the approach provides new independent means of validating protein hits and unbiased identification of polymorphic sequences, sites of alternative splicing and posttranslational modifications. Oral Communications