Biomedical subjects

Jacques Colinge

Publications and source records attributed to Jacques Colinge.

3 recordsLinked to original sources

PubMed2003

OLAV: towards high-throughput tandem mass spectrometry data identification.

Mass spectrometry combined with database searching has become the preferred method for identifying proteins in proteomics projects. Proteins are digested by one or several enzymes to obtain peptides, which are analyzed by mass spectrometry. We introduce a new family of scoring schemes, named OLAV, aimed at identifying peptides in a database from their tandem mass spectra. OLAV scoring schemes are based on signal detection theory, and exploit mass spectrometry information more extensively than previously existing schemes. We also introduce a new concept of structural matching that uses pattern detection methods to better separate true from false positives. We show the superiority of OLAV scoring schemes compared to MASCOT, a widely used identification program. We believe that this work introduces a new way of designing scoring schemes that are especially adapted to high-throughput projects such as GeneProt large-scale human plasma project, where it is impractical to check all identifications manually.

Bayes Theorem↗

PubMed2003

Improved peptide charge state assignment.

We present a new approach capable of assigning charge states to peptides based on both their intact mass spectrum and their fragmentation mass spectrum. More specifically, our approach aims at fully exploiting available information to improve correct charge assignment rate. This is achieved by using information provided by the fragmentation spectrum extensively. For low-resolution spectra, charge assignment based on fragmentation mass spectrum is better than charge assignment based on intact peptide signal only. We introduce two methods that allow to integrate information contributing to successful peptide charge state assignment. We demonstrate the performance of our algorithms on large ion trap data sets. The application of these algorithms to large-scale proteomics projects can save significant computation time and have a positive impact on identification false positive rates.

Algorithms↗

PubMed2002

MIG--differential gene expression in mouse brain endothelial cells.

Different diseases of the CNS are associated with blood-brain barrier (BBB) damage and mononuclear cell infiltration. In order to study genes that may play a role in endothelial cell regulation in inflammatory CNS diseases, we performed differential gene expression (DGE) analysis using a mouse brain endothelial cell line. We found that interferon-gamma (IFNgamma)-induced monokine (MIG), a chemokine that plays a role in T lymphocyte and monocyte chemoattraction, is highly expressed in the presence of inflammatory cytokines. We show that MIG, produced by brain endothelial cells in vitro, is biologically active in attractingT lymphocytes and that it is possible to interfere with this mechanism of action using anti-MIG antibodies. We suggest that blocking MIG may be beneficial in CNS inflammation. We detected constitutive expression of the MIG receptor, CXCR3, on the surface of the endothelial cells and therefore hypothesize that it plays a role in maintaining the cytokine gradient at the region of CNS inflammation.

Animals↗