Design of Experiments (DoE) has been widely used for fermentation development and optimization. However, traditional statistical DoE methods can be time-consuming and cost prohibitive for complex systems with over 20 factors due to the large number of experimental runs.
A novel methodology based on system theories and principles, rather than traditional statistics-based DOE, was used to optimize a complex fermentation process with 66 parameters (including base media, feed solutions, and fermentation operation parameters). A 4-phase study including14 fermentation runs was executed. The optimized conditions resulted in a 3-fold increase in the target protein titer (from 4.2 g/L at initial baseline conditions to 13.2 g/L
As a comparison, an independent study using statistical DoE methodology was executed to optimize the target protein productivity by evaluating base media and process conditions. This DoE approach, which included only 6 factors and a total of 20 runs, increased the target protein yield to 10 g/L from the baseline conditions.
Overall, even with relatively fewer runs, the novel methodology enabled optimization of a complex system with 66 parameters and achieved higher titer compared to the optimization performed using the traditional statistics DoE method.