Nonribosomal peptides are molecules produced by microorganisms and displaying a broad spectrum of biological activities and pharmaceutical applications. They can harbor anti-microbial, immuno-modulating or anti-tumor activities. These peptides are synthesized by huge multi-enzymatic complexes, called NonRibosomal Peptide Synthetases. Two main structural traits distinguish these peptides from the ribosomally synthesized ones : first, their primary structure is not always linear but is often cyclic (partially or totally), branched or poly-cyclic, and second, the diversity of monomers incorporated into nonribosomal peptides extends far beyond the 20 proteogenic amino acids residues. We have developed Norine, the first public resource entirely dedicated to nonribosomal peptides. Norine currently contains more than 1 000 peptides, modeled by non-oriented labeled graphs, and computational tools allowing their analysis, such as monomer composition comparison, structural pattern matching or similarity search. Statistical analysis of Norine data highlighted interesting biological properties such as a speci?c monomer composition depending on the biological activity, that led us to develop a tool for helping the prediction of peptide activity from its monomeric composition. In three years, Norine became the international resource for nonribosomal peptides.
Directeur de Thèse : Gregory Kucherov Co-encadrant : Maude Pupin Rapporteurs : Marie-Dominique Devignes Pierre Cornelis Membres : Bernard Wathelet Daslav Hranueli Valérie Leclère Philippe Jacques
Thèse de l'équipe Bonsai soutenue le 8 septembre 2009