Using predictive modeling to optimize bioreactor performance
Using a predictive modeling approach can help increase operator control over bioprocessing, which in turn can increase bioreactor production.
As the biopharmaceutical industry moves toward automated processes, technologies that allow biomanufacturers to predict outcomes and solve problems are becoming increasingly useful. Predictive modeling is one such technology that can be used to optimize bioreactor performance.
The issue of controlling bioreactors has become increasingly important due to the inherent difficulty in controlling the behavior of living cells used in cell culture systems (1). Despite a well-designed control scheme, a bioreactor can still exhibit poor performance. As such, an approach that includes predictive modeling can be an advantage in developing process controls for bioreactor performance.
Predictive modeling approach
One of the benefits of using predictive models is that these models allow for a data-driven approach, which can help speed up process development, says Tania Pereira Chilima, chief technology officer at Univercells Technologies. Chilima explains that these models allow users to gain a deeper understanding of the processes, which over time helps to limit the development resources required.
“This approach can be used in both batch and continuous processing and can perhaps be even more useful in continuous processing due to the increased number of variables to understand and control (e.g. Getting all of this data in real time allows for a faster learning curve and helps ensure a repeatable process,” says Chilima.
Edita Botonjic-Sehic, director of analytics at Pall, notes that predictive modeling combines tools and technology “to collect real-time process data and turn it into meaningful long-term insights.” To do this, predictive modeling technology uses mathematical algorithms to transform data streams extracted from batch or continuous processes into meaningful and targeted hypotheses.
“You can think of it like Amazon’s algorithm that uses search and purchase data to make certain assumptions and suggestions. Over time, that information gets richer,” says Botonjic-Sehic. “When we adopt this approach to bioprocessing, it predicts the process conditions, such as the outcome of the glucose feed rate during the process.”
Although predictive modeling can be used in both batch and continuous processes, however, it is currently more difficult to successfully apply it to continuous processing, explains Botonjic-Sehic. Companies are striving to fill the gaps with batch and continuous predictive modeling solutions that drive better results.
Upstream challenges for modeling to be resolved
The upstream processing step can be difficult. On the one hand, upstream proteins are not pure and must be differentiated, and on top of that, there are multiple attributes and parameters that must be monitored and controlled with analytical devices to maintain process conditions, Botonjic points out. -Sehic. These parameters include temperature, pressure, dissolved oxygen, pH, and critical quality attributes (CQA), all of which should be part of the model.
“Multi-variant statistical tools are needed to find the right models due to the complexity and attribute interactions in the bioreactor; the need for the right tools to measure attributes in real time is a need, and, even when data is acquired, there remains a challenge in translating that data into meaningful information that is ultimately used for prediction,” says Botonjic-Sehic.
Analytical tools such as Raman spectroscopy are commonly used along with other complex spectroscopic techniques to help measure analytes and predict information using predetermined models. Botonjic-Sehic says the industry is pushing to find new solutions to support better predictive modeling.
“The ability to monitor all AQCs of cell growth using predictive models would be useful not only to streamline glucose, nutrient and metabolite utilization, but also to significantly reduce overall process waste and time. of development. The ability to predict cell viability and titer concentration is also useful for targeting the right amount of protein to grow [into downstream processing]which is a much purer product,” says Botonjic-Sehic.
Additionally, many advances have been made in the area of bioreactors, and Pall is also working on its own solutions. “The key here is small-scale bioreactor modeling. The ability to demonstrate feasibility early in processes will be key to reducing costs and increasing accessibility. We need cross-functional solutions that can measure, monitor, predict and control processes,” says Botonjic-Sehic.
Upstream predictive modeling approaches can also impact design
innovations or development of accessories moving forward.
Chilima notes that the industry has certainly seen an increase in the number of process analytics technologies (PATs) being developed, for example. PAT enables automated real-time data collection. “Some of these tools also allow different types of data to be collected with a single probe, which is a real advantage. Lately, we have also seen significant progress in software development with a number of tools that centralize data collection and analysis to facilitate the development of predictive models, and even digital twins,” says Chilima.
“We expect to see different types of models and approaches taking into account a multitude of processes,” adds Botonjic-Sehic. She notes that more work is underway in the space to deliver an enabling, standardized platform approach to predictive modeling of the discovery phase at commercial scale.
Another innovative tool that can enable predictive approaches in upstream processing is the microbioreactor. This is an area Univercells has focused on because microbioreactors can be used to collect an immense amount of data without significant resources, says Chilima. “When the operation of these bioreactors is automated, the costs of developing reliable predictive models can decrease significantly. Removing the cost barrier will facilitate more innovation in the space and continue to evolve the spread of modalities in the industry. It’s an exciting time for development and manufacturing,” remarks Chilima.
“It’s a great place for Pall,” Botonjic-Sehic said. The company continues to expand its small-scale and large-scale offerings. “We recognize that the industry needs flexible models that streamline development and enable efficient and reliable technology transfer. The idea is to align this flexibility with robust modeling and predictive maintenance functionality from the microbioreactor. »
1. S. Ramaswamy, TJ Cutright and HK Qammar, Process biochemistry 40(8) 2763–2770 (2005).
About the Author
Feliza Mirasol is the scientific editor of BioPharm International.