Découverte de motifs relationnels en bioinformatique: application à la prédiction de ponts disulfures

Ingrid Jacquemin 1
1 SYMBIOSE - Biological systems and models, bioinformatics and sequences
IRISA - Institut de Recherche en Informatique et Systèmes Aléatoires, Inria Rennes – Bretagne Atlantique
Abstract : The main properties of proteins may be determined from their three-dimensional structure. Bridging this ``knowledge gap'' requires automatic methods capable of inferring folds from sequences. Our contribution to this ambitious goal addresses the prediction of particular interactions inside these macro-molecules. The tertiary folds of native proteins are defined by a large number of weak interactions. Proteins are also stabilized covalently by disulfide bonds formed by a pair of cysteine residues in the folded state. We focus on the prediction of these disulfide bonds formed by a pair of cystein residues in the folded state.
We propose to tackle with the problem of prediction of interactions between cysteins in the framework, in the first step, of grammatical inference, especially control sets introduced by Y. Takada and, in the second step, of inductive logic programming. A lot of experiments show that grammatical inference algorithms have to be improve to be able to work on protein sequences. On the over hand, we obtain explicit rules of prediction of oxidized state of cysteins which can be easily interpreted by biologists. We propose an heuristic induction algorithm which can be use for different problems. We obtain several rules and we can see that with only 17 rules, more than 90\% of the two data bases we use are covered, which is a very good result given the simplicity of the characterization.
Complete list of metadatas

https://tel.archives-ouvertes.fr/tel-00185499
Contributor : Anne Jaigu <>
Submitted on : Tuesday, November 6, 2007 - 11:42:15 AM
Last modification on : Friday, November 16, 2018 - 1:23:34 AM
Long-term archiving on : Monday, September 24, 2012 - 2:51:31 PM

Identifiers

  • HAL Id : tel-00185499, version 1

Citation

Ingrid Jacquemin. Découverte de motifs relationnels en bioinformatique: application à la prédiction de ponts disulfures. Biochimie [q-bio.BM]. Université Rennes 1, 2005. Français. ⟨tel-00185499⟩

Share

Metrics

Record views

1163

Files downloads

144