17th International Mass Spectrometry Conference :: Prague, 2006
> Go to contents (site navigation)
|Presentation date:||Mon, Aug 28, 2006|
|Presentation time:||12:10 – 12:30|
Miren J. Omaetxebarria2, Felix Elortza3, Jesus M. Arizmendi2, Ole N. Jensen4, Rune Matthiesen11 CIC bioGUNE, DERIO, Spain
Correspondence address: Rune Matthiesen, CIC bioGUNE, Bionformatics, Parque Tecnologico Edificio 801 A, DERIO, 48160 Spain.
Keywords: Computational Methods; Computer Program; Data Analysis; Mass Spectrometry.
Novel aspect: A new combinatorial algorithm for the identification of GPI anchored peptides in MS/MS data.
High throughput GPI-anchored protein identification is nowadays based on the concept of modification-specific proteomics, where plasma membrane preparation, PI-PLC/PLD treatment, LC-MS/MS and computational sequence analysis are combined. Computational sequence analysis is based on the use of prediction tools that consist of algorithms that search for specific features of GPI-anchored pro-proteins as it is the presence of a N-terminal signal sequence, presence of a hydrophobic C-terminal sequence and the assignment of the omega-site to small residues. Experimental verification of the existence of a GPI-anchored protein is a tedious task and as a consequence less than 10% of the proteins that are described as GPI-APs in the Swissprot database have been experimentally verified. GPI-anchors described to date contain a core structure which is well conserved among species. The presence of a non-acetylated glucosamine residue in the core structure is a unique feature that is not generated by N-linked or O-linked glycans and therefore it is a candidate diagnostic signal for GPI-APs. Detection by mass spectrometry of specific ions of GPI-APs could potentially be used for precursor ion scanning experiments to detect with high specificity and sensitivity the presence of GPI-anchored peptides in complex samples. We described herein a new combinatorial algorithm for identifying GPI anchored peptides in MS/MS spectra. To date no other MS/MS search engine are able to identify GPI anchored peptides. GPI anchored peptides are a challenge to identify due to the heterogeneity of the core glycan structure and the semi tryptic nature of the GPI-anchored C-terminal peptide. These two factors give an expanded search space which requires a new combinatorial algorithm and scoring method. The expanded search space leads to increased computational overhead. The search space is therefore restricted by a GPI anchor predictor based on machine learning so only the most likely cleavage and GPI anchor sites are considered. The presented combinatorial algorithm is an expansion of the combinatorial algorithm in VEMS.1,2
1. R. Matthiesen, J. Proteome Res. 4, 2338 (2005).
2. R. Matthiesen, Proteomics 4, 2583 (2004).