Prediction of New Bioactive Molecules using a Bayesian Belief Network

Abstract : Natural products and synthetic compounds are a valuable source of new small molecules leading to novel drugs to cure diseases. However identifying new biologically active small molecules is still a challenge. In this paper, we introduce a new activity prediction approach using Bayesian belief network for classification (BBNC). The roots of the network are the fragments composing a compound. The leaves are, on one side, the activities to predict and, on another side, the unknown compound. The activities are represented by sets of known compounds, and sets of inactive compounds are also used. We calculated a similarity between an unknown compound and each activity class. The more similar activity is assigned to the unknown compound. We applied this new approach on eight well-known data sets extracted from the literature and compared its performance to three classical machine learning algorithms. Experiments showed that BBNC provides interesting prediction rates (from 79% accuracy for high diverse data sets to 99% for low diverse ones) with a short time calculation. Experiments also showed that BBNC is particularly effective for homogeneous data sets but has been found to perform less well with structurally heterogeneous sets. However, it is important to stress that we believe that using several approaches whenever possible for activity prediction can often give a broader understanding of the data than using only one approach alone. Thus, BBNC is a useful addition to the computational chemist's toolbox. ■ INTRODUCTION Due to the similar property principle, 1 structurally similar compounds are expected to exhibit similar properties and similar biological activities. This principle is exploited for in silico discovery of new drugs with the emergence of an activity prediction technology based on chemical structures. A variety of computational approaches for target or activity prediction were published over the past several years. For example, quantitative structure−activity relationship (QSAR) 2−5 was established on the hypothesis that compounds with similar physicochemical properties and/or structure share similar activities. The effectiveness of a QSAR analysis relies both on selecting the relevant descriptors for modeling the biological activity of interest and on the choice of a good quantitative model that maps the compound descriptors to chemical property or biological activity by means of statistical techniques. In similarity searching strategies, an unknown compound (the target) is compared to a set of compounds with known activities. The activity for which the compounds are the most similar to the target is assigned to it. Binary kernel discrimination (BKD), 6,7 naı̈ ve Bayesian classifier (NBC), 8−11
Type de document :
Article dans une revue
Journal of Chemical Information and Modeling, American Chemical Society, 2014, 54 (1), pp.30-36. 〈〉. 〈10.1021/ci4004909〉
Liste complète des métadonnées

Littérature citée [12 références]  Voir  Masquer  Télécharger
Contributeur : Maude Pupin <>
Soumis le : vendredi 5 décembre 2014 - 10:31:54
Dernière modification le : vendredi 8 janvier 2016 - 01:06:55
Document(s) archivé(s) le : lundi 9 mars 2015 - 06:01:43


Fichiers éditeurs autorisés sur une archive ouverte



Ammar Abdo, Valérie Leclère, Philippe Jacques, Naomie Salim, Maude Pupin. Prediction of New Bioactive Molecules using a Bayesian Belief Network. Journal of Chemical Information and Modeling, American Chemical Society, 2014, 54 (1), pp.30-36. 〈〉. 〈10.1021/ci4004909〉. 〈hal-01090611〉



Consultations de
la notice


Téléchargements du document