A new fingerprint to predict nonribosomal peptides activity.

Ammar Hasan Abdo (1, 2) , Ségolène Caboche (1, 2) , Valérie Leclère (3) , Philippe Jacques (3) , Maude Pupin (1, 2)

1 LIFL - Laboratoire d'Informatique Fondamentale de Lille
2 BONSAI - Bioinformatics and Sequence Analysis
3 PROBIOGEM - Laboratoire des Procédés Biologiques Génie Enzymatique et Microbien

Bacteria and fungi use a set of enzymes called nonribosomal peptide synthetases to provide a wide range of natural peptides displaying structural and biological diversity. So, nonribosomal peptides (NRPs) are the basis for some efficient drugs. While discovering new NRPs is very desirable, the process of identifying their biological activity to be used as drugs is a challenge. In this paper, we present a novel peptide fingerprint based on monomer composition (MCFP) of NRPs. MCFP is a novel method for obtaining a representative description of NRP structures from their monomer composition in fingerprint form. Experiments with Norine NRPs database and MCFP show high prediction accuracy (>93 %). Also a high recall rate (>82 %) is obtained when MCFP is used for screening NRPs database. From this study it appears that our fingerprint, built from monomer composition, allows an effective screening and prediction of biological activities of NRPs database.

Mots clés: Nonribosomal peptides; Target Prediction; Similarity searching; Drug discovery

Domaines: Bio-informatique [q-bio.QM]; Bio-Informatique, Biologie Systémique [q-bio.QM]; Chemo-informatique

Ammar Hasan Abdo, Ségolène Caboche, Valérie Leclère, Philippe Jacques, Maude Pupin. A new fingerprint to predict nonribosomal peptides activity.. Journal of Computer-Aided Molecular Design, 2012, 26 (10), pp.1187-94. ⟨10.1007/s10822-012-9608-4⟩. ⟨hal-00750002⟩