Section: New Results

Nonribosomal peptides

  • Monomeric structure. The algorithm that identifies the monomeric structure of a polymer from its chemical structure has been finished and named s2m. It is based on a double index: A partial index constructed on the monomer database that uses a markovian model to speed up the search time ; and an index constructed on the fly on the studied polymer. This strategy was originally developed for nonribosomal peptides, but can be applied to any polymer.

  • Florine: Nonribosomal peptide synthetase annotations. Florine [2] is a workflow dedicated to the discovery of new nonribosomal peptide synthetases. It describes sequential steps starting from DNA sequences leading to the design of candidate bioactive peptides. It is a useful tool for new drug discovery because it can be applied whatever the producing micro-organisms as it takes into account the enzymatic specificities related to each genus. This work was performed in collaboration with members of EPI Orpailleur (CRI Nancy Grand Est), Marie-Dominique Devignes and Malika Smaïl-Tabbone.

  • Activity prediction of small molecules. Bayesian Belief Network was used for the first time to classify compounds according to their biological activity [1] . This method was applied on nonribosomal peptides and gave promising results on predicting their activity.