| Popis: |
Shaken Erlenmeyer flasks are a commonly used cultivation system in biotechnology and are regularly employed in early-stage process development (e.g., inoculum production, media development, and strain characterization). However, they are mostly used as black box systems without the sensoring capabilities a stirred tank bioreactor can provide. Several measurement systems have been developed to overcome this issue, but there is still a lack of comparability or a uniform, automated approach to using online data for process characterization in shake flasks. To overcome this, we compared online backscattered light, dissolved oxygen, and pH data for plant, animal, E. coli, and S. cerevisiae cultivations using the PreSens SFR vario. With these data, key performance indicators (KPIs), such as the specific growth rate and the cell-specific oxygen demand, were evaluated automatically. For algorithm validation, manually calculated KPIs based on offline biomass data, online data and algorithm-based automatically calculated KPIs using online data were compared. The developed algorithm is based on Python and searches for the exponential phase in the corresponding online signal. The exponential fit set by the algorithm and the observed signal were compared and the fit optimized so that the root-mean-square error was as low as possible. With this technique, an accurate estimation of the growth rate and further calculation of the cell-specific oxygen demand can be performed using either the oxygen uptake rate or a biomass estimation based on the backscattered light signal. This enables the comparison and evaluation of different media, strains, and process conditions in a standardized cost-effective and automated manner, reducing human effort and errors. |
