Digital Penicillin Production

For thousands of years, micro-organisms have been used to facilitate chemical reactions - in beer brewing, for example. However, biochemical processes are incredibly complex, with a multitude of reactions taking place simultaneously and influencing one another. There are countless parameters that play a role, not all of which can be directly measured.

Despite the difficulties involved, researchers at TU Wien are now working on examining these processes in detail. Now, in cooperation with pharmaceutical manufacturer Sandoz, TU Wien has succeeded in analysing and comprehensively replicating a penicillin production process using a computer model. This process has even enabled researchers to determine parameters that cannot be measured directly. Sandoz is now making use of these findings to keep a full overview of the processes in the bioreactor at all times, ensuring optimum quality.

Black box replaced by in-depth knowledge

Many chemical reactions are easy to understand: if hydrogen is burned with oxygen, water is produced - in a clearly predictable manner and in a volume that can be precisely calculated beforehand. But how can you calculate how quickly a fungus will grow and proliferate under the constantly changing conditions in a bioreactor?

"For a long time, processes like this were seen as a 'black box' that cannot be understood and that can only be effectively exploited with a lot of experience," says Prof. Christoph Herwig, who leads the research group for bioprocess technology at TU Wien's Institute of Chemical, Environmental and Bioscience Engineering. "Our approach is somewhat different: we want to analyse the chemical processes in a bioreactor in detail and determine the equations that describe these processes." The aim is to produce a mathematical model that accurately replicates these processes within the bioreactor.

"Many parameters that are vital to the process simply cannot be directly measured, such as the growth rate of the micro-organisms," explains Julian Kager, who is working with Sandoz GmbH as part of his dissertation. "This is precisely why a comprehensive mathematical model is so useful: we use accessible data from the production process in real time, such as the concentration of various substances in the bioreactor, and use our computer model to calculate the most probable state of the process." The parameters that can't be measured can therefore be calculated.

The model information can be used to optimise the nutrient supply to the cultivated cells while the process is ongoing.

The system of equations used to mathematically describe the bioprocess is as complex and multifaceted as the process itself. "The system of equations describes a non-linear dynamic system. Even the smallest variations in the starting conditions can have a huge impact," explains Kager. "This means it's not really possible to simply work out a solution by hand; instead, relatively elaborate computer simulations are required to obtain the needed results."

The process model and algorithms developed at TU Wien are now being used by Sandoz GmbH for its penicillin production process. "We are very pleased that our basic research has been adopted for use in industry so quickly and that our approach of biochemical modelling is now being used to facilitate the automated control of pharmaceutical production processes," says Julian Kager.

Julian Kagera, Christoph Herwigab, Ines Viktoria Stelzercd.
State estimation for a penicillin fed-batch process combining particle filtering methods with online and time delayed offline measurements.
Chemical Engineering Science, Volume 177, 23 February 2018, Pages 234-244. doi: 10.1016/j.ces.2017.11.049.

Most Popular Now

NHS Staff Punished as 500,000 Rely on Wh…

WhatsApp, Facebook Messenger and other unauthorised instant messaging (IM) apps are being used by approximately 500,000 NHS staff at work, as a growing number turn to consumer tools to communicate...

Call for Abstracts: European Telemedicin…

27 - 29 May 2018, Sitges, Barcelona, Spain. The European Telemedicine Conference 2018 (ETC18) is an interdisciplinary forum for healthcare professionals, directors, managers, and researchers with the intent of bringing together...

conhIT 2018: The stage is Set for Dialog…

17 - 19 April 2018, Berlin, Germany. Finding out about and supporting all aspects of the digital transformation of the healthcare system: that is what this year's conhIT, Europe's largest event...

Smartphone 'Scores' can Help Doctors Tra…

Parkinson's disease, a progressive brain disorder, is often tough to treat effectively because symptoms, such as tremors and walking difficulties, can vary dramatically over a period of days, or even...

Focus on the Digital Transformation - A …

17 - 19 April 2018, Berlin, Germany. How is the digitalisation of the healthcare system affecting the relationship between patients and doctors? What new innovations and solutions does the health IT...

Portable Device Detects Severe Stroke in…

A new device worn like a visor can detect emergent large-vessel occlusion in patients with suspected stroke with 92 percent accuracy, report clinical investigators at the Medical University of South...

Imitation is the Most Sincere Form of Fl…

For every two mobile apps released, one is a clone of an existing app. However, new research published in the INFORMS journal Information Systems Research shows the success of the...

Merck Partners with Medisafe to Help Imp…

Merck, a leading science and technology company, today announced a new collaboration with US-based Medisafe to help its cardiometabolic patients better manage medication intake and adhere to prescribed treatment regimens...

Philips Research-led Big Data Consortium…

Royal Philips (NYSE: PHG, AEX: PHIA), a global leader in health technology, together with its consortium partners, today announced that it has received funding from the EU's Horizon 2020 program...

Smartphone App Performs Better than Trad…

A smartphone application using the phone's camera function performed better than traditional physical examination to assess blood flow in a wrist artery for patients undergoing coronary angiography, according to a...

Deep Learning Transforms Smartphone Micr…

Researchers at the UCLA Samueli School of Engineering have demonstrated that deep learning, a powerful form of artificial intelligence, can discern and enhance microscopic details in photos taken by smartphones...