Experimental design and Bayesian networks for enhancement of delta-endotoxin production by Bacillus thuringiensis

Abstract

Bacillus thuringiensis (Bt) is a Gram-positive bacterium. The entomopathogenic activity of Bt is related to the existence of the crystal consisting of protoxins, also called delta-endotoxins. In order to optimize and explain the production of delta-endotoxins of Bacillus thuringiensis kurstaki, we studied seven medium components: soybean meal, starch, KH 2 PO 4 , K 2 HPO 4 , FeSO 4 , MnSO 4 , and MgSO 4 and their relationships with the concentration of delta-endotoxins using an experimental design (Plackett-Burman design) and Bayesian networks modelling. The effects of the ingredients of the culture medium on delta-endotoxins production were estimated. The developed model showed that different medium components are important for the Bacillus thuringiensis fermentation. The most important factors infl uenced the production of delta-endotoxins are FeSO 4 , K 2 HPO 4 , starch and soybean meal. Indeed, it was found that soybean meal, K 2 HPO 4 , KH 2 PO 4 and starch also showed positive effect on the delta-endotoxins production. However, FeSO 4 and MnSO 4 expressed opposite effect. The developed model, based on Bayesian techniques, can automatically learn emerging models in data to serve in the prediction of delta-endotoxins concentrations. The constructed model in the present study implies that experimental design (Plackett-Burman design) joined with Bayesian networks method could be used for identifi cation of effect variables on delta-endotoxins variation.

Evaluation of Variability in Tunisian Olea europaea L. Accessions using Morphological Characters and Computational Approaches

“…from artificial intelligence studies and constitute one of the most coherent techniques for the acquisition and the modelling of complex systems. They have been applied to a large range of problems, and eventually in biology (Ennouri et al, 2015. Since the data are discontinuous and experimental data are limited, and it is well known that the application of BN requires a lot of data for the learning and testing procedures, our proposed methodology includes the following three different stages for building model: Data normalization and Construction of Bayesian Network.…”