Computational intelligence applied to monitoring and control of bioreactors during the cultivation of wild and recombinant microorganisms to produce penicillin G acylase

Abstract

The focus of this project is the application of modern computational intelligence techniques to the monitoring, on-line state inference and control of stirred-aerated tank bioreactors. The project aims at developing and implementing computational tools, necessary to optimize the production of the enzyme penicillin G acylase (PGA) by different wild and recombinant microorganisms, thus providing essential information for the industrial scale-up of the process.The proponent's research group has been working on bioreactor automation for years, an interdisciplinary research field that has a close cooperation with laboratories of Biochemical Engineering from the Department of Chemical Engineering, of Molecular Biology and Biochemistry from the Department of Physiological Sciences, and of Artificial Intelligence and Automation from the Computation Department, all from the Universidade Federal de São Carlos. This project is intended to support these activities, focusing an enzyme of great interest for the pharmaceutical industry, penicillin G acylase (PGA), which is extensively used in industry, for the hydrolysis of microbial penicillins.Hence, this project unfolds into two directions: using the production of PGA as a case study for applying and testing different computational intelligence techniques, always using data from actual experiments, gathered from automated bench-scale 5-L fermentors; and supporting the optimization of the production process, integrated to downstream operations, while providing information about the metabolic profile of the studied microorganisms. Having the necessary framework for on-line data acquisition, monitoring and control of the bioreactors, different computational routines for non-linear systems will be studied and compared, focusing on signal processing, state inference, fault detection and optimization of the cultivations. Thus, besides advancing the know-how in the production of an important biocatalyst for the pharmaceutical industry, the expertise of our research group concerning the use of computational intelligence for bioreactors automation will be consolidated, in an area that has been experiencing a worldwide explosive growth. (AU)