Computational tool for automated large-scale GPIomic analysis

Juan Clemente Aguilar Bonavides, University of Texas at El Paso

Abstract

Liquid chromatography-tandem mass spectrometry (LC-MS/MS or MS/MS) is the most efficient tool today for the identification of glycosylphosphatidylinositol (GPI) molecules. The amount of data produced in each MS/MS experiment is a major bottleneck in high-throughput GPIomic (the entire collection of free and protein-linked GPIs) projects. Efficient computational tools can significantly reduce the amount of time analyzing MS/MS data; however, at present the automatic interpretation of these data to annotate GPI structures is absent. We propose a library-based tool to identify GPI structures by matching fragment peaks in the spectra with data derived from a theoretical database of GPI structures that we have developed. Currently, our scoring method produces an overlap between the scores of correct and incorrect structures identified since a number of isobaric structures exist within the database. Thus, to ensure that most of the identifications are true positive, a scoring system with better specificity must be applied. Considering the success of the peptide identification approach and the new methodologies for glycan and lipid identification, we expect that the development of a new method that combines techniques can be developed in combination with our existing tool.

Subject Area

Bioinformatics

Recommended Citation

Aguilar Bonavides, Juan Clemente, "Computational tool for automated large-scale GPIomic analysis" (2011). ETD Collection for University of Texas, El Paso. AAI1494323.
https://scholarworks.utep.edu/dissertations/AAI1494323

Share

COinS