Seed Train Tool
Optimizing the Seed Train
Therapeutic proteins, vaccines, and monoclonal antibodies—the market for biopharmaceuticals has become a key growth driver for the industry. However, production conditions in bioreactors are highly complex, and manufacturers must comply with strict regulatory quality requirements. The biotechnology research conducted by Professor Björn Frahm and his team provides valuable support in addressing a central challenge: How can the manufacturing process—known as the seed train—be designed to reliably generate the trillions of cells required for production while ensuring quality, stability, and efficiency?
Cell Culture Technology as the Basis for Biopharmaceutical Production
Biopharmaceuticals such as monoclonal antibodies for cancer diagnosis and therapy, the treatment of autoimmune diseases, or use in rapid diagnostic tests (e.g., pregnancy or antigen tests) are produced using cell culture technology. Monoclonal antibodies are also applied in laboratory medicine and protein analysis research. Additional applications of cell cultures include the production of vaccines and therapeutic proteins, such as blood coagulation factor VIII for the treatment of hemophilia.
Currently, more than 350 biopharmaceuticals are approved worldwide, with approximately 30–40 new active ingredients added each year. According to market research reports, global sales of biopharmaceutical products are expected to exceed US$500 billion by 2025. The technology required for cell culture production—including specialized equipment—is growing at an annual rate of approximately 12% and was valued at more than US$20 billion last year. Advances in research, global population growth, and improved healthcare in many countries are expected to drive further expansion of the market.
Mammalian cells are predominantly used in biopharmaceutical cell culture technology because—unlike microbial expression systems—they are capable of producing complex proteins with correct three-dimensional folding and human-like post-translational modifications.
Further information on cell culture technology and how it clearly differs from animal testing can be found here.
The Challenges of Cell Culture Propagation
Cells and their properties may change over time, potentially leading to reduced effectiveness. In cell culture technology, this issue is addressed by cryopreserving fresh cells so that they can be regularly thawed and expanded when needed. Repeated thawing from a frozen cell bank ensures stable and reproducible production of high-quality products.
However, this approach requires that approximately 5 trillion cells be newly cultivated each time a 10,000-liter bioreactor is started. This process is referred to as the seed train (or inoculation train).
The cell number is increased in a stepwise process, transferring them usually into larger cultivation systems. The production bioreactor (also referred to as main stage) is inoculated out of the largest seed train scale. A cell culture seed train lasts for a significant period of time, e.g. 3 – 4 weeks and generates corresponding costs.
The seed train is crucial for production from the very beginning: cells must be propagated under optimal cultivation conditions, as both cell growth and product formation are significantly influenced by their environment up to and including the production bioreactor. Numerous process parameters can be adjusted to design the seed train and influence its progression—resulting in multidimensional design and optimization challenges.
Seed Train Research at OWL University of Applied Sciences and Arts
At OWL University of Applied Sciences and Arts in Lemgo, Germany, Professor Björn Frahm’s biotechnology group conducts research on cell culture seed trains. This work has led to the development of the Seed Train Tool—a software solution that digitally represents a wide range of seed train scenarios across different cell lines, products, and companies.
“The Seed Train Tool enables complex cell proliferation processes to be digitally mapped in a way that opens up diverse applications for industry and research,” explains Björn Frahm.
The resulting digital twin enables the analysis and optimization of existing seed trains, as well as monitoring—i.e., the prediction and adjustment—of ongoing production processes. Using the digital twin, it is possible to simulate how cells respond to process changes. This enhances product quality and stability while significantly reducing the need for costly experimental trials.
Furthermore, the Seed Train Tool supports the design of new seed trains, the development of seed train protocols, and the dimensioning of associated bioreactor systems. Creating a seed train protocol involves defining the process parameters under which the seed train is carried out—for example, the cell concentration at which progressively larger bioreactors are inoculated, how cells proliferate within each stage, how transfers between scales are performed, and how much nutrient medium—with which specific concentrations of key components—must be added.
In addition, the seed train can be modeled using Gaussian processes and optimized holistically across multiple criteria. These extensions were developed in collaboration with Professor Markus Lange-Hegermann.
Another industrial application addressed by the Seed Train Tool is the transfer of existing seed trains to other production facilities manufacturing the same product.
A further application lies in cell line development, where a suitable origin cell must be selected that combines high product quality, productivity, stability, and rapid growth—while also being compatible with the future seed train process.
The Seed Train Tool has been adopted by Novartis in Austria and has also contributed valuable insights to joint research projects with ExcellGene in Switzerland. A research collaboration with Sandoz is also in place.
Currently, structured approaches are being compiled to translate Seed Train knowledge into practical tools—for example, decision trees, decision rules, and criteria for transferring cells to the next scale. These approaches are analyzed and compared in terms of their respective advantages and disadvantages. The aim is to address challenges faced by many companies worldwide in this field.
In addition, the Seed Train Tool is now available with a simplified interface for routine industrial use, complementing the developer mode designed primarily for research applications.
Uncertainty modeling and statistical methods also play a key role in the development of the Seed Train Tool: What uncertainties arise from sampling, measurement inaccuracies, and simulation? How can prior knowledge—such as expert input and experimental data—be systematically integrated? With what probability can reliable predictions be made?
These workflows are also relevant beyond biopharmaceutical production. Dr. Tanja Hernández Rodríguez, who has conducted extensive research on seed trains and uncertainty modeling for many years, has successfully transferred this expertise across domains.
Dictionary
Cell culture technology is based on established cell lines and is clearly distinct from animal testing. It involves working exclusively with cells that proliferate in culture systems such as bioreactors. These cell lines originate from a single biological sample, often obtained many years or even decades ago. No living animals are required or harmed in routine cell culture processes.
Today, CHO cells are the most commonly used production cells in the biopharmaceutical industry and are responsible for the manufacture of approximately 70–80% of biologics.
Economical production typically takes place in bioreactors with volumes ranging from 1,000 to 20,000 liters. In well-optimized processes, product titers currently reach 5–10 grams per liter.
A 10,000-liter bioreactor inoculated with mammalian cells typically starts with a cell concentration of 200,000–500,000 cells per milliliter. This corresponds to a total of up to 5 × 10¹² cells—i.e., approximately 5 trillion (5,000 billion) cells.