Cell-cultivated (or “lab-grown”) meat offers a sustainable, slaughter-free alternative to traditional animal agriculture by growing animal cells in bioreactors. This emerging field draws from traditional biopharmaceutical manufacturing but requires lower input costs (especially cell culture media) to achieve commodity-scale pricing. At UPSIDE Foods, we developed high-throughput experimental workflows to rapidly test and optimize cell culture media formulations. However, large data sets (2000+ measurements per run) created a new bottleneck, where manual data processing slowed iteration cycles and delayed process decisions.
To overcome this, we set up an Open Database Connection (ODBC), enabling real-time access to key instrument data (e.g., ViCell XR/Blu, Nova Flex 2). We then developed tools using the JMP Scripting Language (JSL) to pull data into JMP and automate our data analysis workflows. This presentation demonstrates the script's functionality, such as built-in data visualizations (Graph Builder), automated trend analysis (Bivariate Fit), and operator-facing tools (Tabulate) to guide process interventions (e.g., feeds, set point adjustments). Our approach also unlocked more advanced calculations requiring numerical integration between data points with conditional relationships. As a case study, I use cell-specific glucose consumption qGluc (pg/(cell*day) to highlight three critical aspects of our methodology that facilitated automation of this notoriously laborious calculation:
- Developing standardized sample naming conventions.
- Automatically organizing and cleaning data using operations like JOIN and SORT.
- Creating formula columns using key functions (e.g., IF, LAG, IS MISSING).
Integrating live data streams with automated analytics in JMP significantly accelerated bioprocess development and reduced decision latency at UPSIDE Foods.
