400,000 in-silico CHO-K1 fed-batch runs with different media concentrations, feeding strategies, and initial seeding densities

A Latin hypercube sampling strategy was applied to randomly generate 1,000 feed media component concentrations and 100 feeding strategies, which combined with 4 base media options resulted in 400,000 unique input combinations. These combinations were modeled by implementing a mechanistic model, which was a combination of a genome-scale metabolic network model with a dynamic flux balance analysis to generate 400,000 in-silico runs for fed-batch culture of CHO-K1 cells at a 250 mL scale. The dataset contains hourly profiles for total, viable, and dead cell densities, as well as concentrations of glucose, 19 amino acids, pyruvate acid and monoclonal antibody (mAb). More details about this dataset can be found in the following publications:
1- A dynamic and generalizable modelling framework based on genome-scale flux balance analysis for Chinese hamster ovary (CHO) fed-batch culture and in-silico dataset generation. Available at SSRN: https://ssrn.com/abstract=5591701 or http://dx.doi.org/10.2139/ssrn.5591701
2- 400,000 in-Silico CHO-K1 fed-batch runs across varying media concentrations, feeding strategies and initial seeding densities
This research was supported under the Australian Research Council’s Industrial Transformation Research Program (ITRP) funding scheme (project number IH210100051). The ARC Digital Bioprocess Development Hub is a collaboration between The University of Melbourne, University of Technology Sydney, RMIT University, CSL Innovation Pty Ltd, Cytiva (Global Life Science Solutions Australia Pty Ltd) and Patheon Biologics Australia Pty Ltd.