Modeling of an industrial process of pleuromutilin fermentation using feed-forward neural networks

Biological and chemical diversity go hand in hand: Basidiomycota as source of new pharmaceuticals and agrochemicals

The Basidiomycota constitutes the second largest higher taxonomic group of the Fungi after the Ascomycota and comprises over 30.000 species. Mycelial cultures of Basidiomycota have already been studied since the 1950s for production of antibiotics and other beneficial secondary metabolites. Despite the fact that unique and selective compounds like pleuromutilin were obtained early on, it took several decades more until they were subjected to a systematic screening for antimicrobial and anticancer activities. These efforts led to the discovery of the strobilurins and several hundreds of further compounds that mainly constitute terpenoids. In parallel the traditional medicinal mushrooms of Asia were also studied intensively for metabolite production, aimed at finding new therapeutic agents for treatment of various diseases including metabolic disorders and the central nervous system. While the evaluation of this organism group has in general been more tedious as compared to the Ascomycota, the chances to discover new metabolites and to develop them further to candidates for drugs, agrochemicals and other products for the Life Science industry have substantially increased over the past decade. This is owing to the revolutionary developments in -OMICS techniques, bioinformatics, analytical chemistry and biotechnological process technology, which are steadily being developed further. On the other hand, the new developments in polythetic fungal taxonomy now also allow a more concise selection of previously untapped organisms. The current review is dedicated to summarize the state of the art and to give an outlook to further developments.

Multiple linear regression and artificial neural networks for delta-endotoxin and protease yields modelling of Bacillus thuringiensis

The aim of the present work was to develop a model that supplies accurate predictions of the yields of delta-endotoxins and proteases produced by B. thuringiensis var. kurstaki HD-1. Using available medium ingredients as variables, a mathematical method, based on Plackett-Burman design (PB), was employed to analyze and compare data generated by the Bootstrap method and processed by multiple linear regressions (MLR) and artificial neural networks (ANN) including multilayer perceptron (MLP) and radial basis function (RBF) models. The predictive ability of these models was evaluated by comparison of output data through the determination of coefficient (R ²) and mean square error (MSE) values. The results demonstrate that the prediction of the yields of delta-endotoxin and protease was more accurate by ANN technique (87 and 89% for delta-endotoxin and protease determination coefficients, respectively) when compared with MLR method (73.1 and 77.2% for delta-endotoxin and protease determination coefficients, respectively), suggesting that the proposed ANNs, especially MLP, is a suitable new approach for determining yields of bacterial products that allow us to make more appropriate predictions in a shorter time and with less engineering effort.