ABSTRACT
Background Arthrospira platensis (commonly known as spirulina) is a promising new platform for low-cost manufacturing of biopharmaceuticals. However, full realization of the platform’s potential will depend on achieving both high growth rates of spirulina and high expression of therapeutic proteins.
Objective We aimed to optimize culture conditions for the spirulina-based production of therapeutic proteins.
Methods We used a machine learning approach called Bayesian black-box optimization to iteratively guide experiments in 96 photobioreactors that explored the relationship between production outcomes and 17 environmental variables such as pH, temperature, and light intensity.
Results Over 16 rounds of experiments, we identified key variable adjustments that approximately doubled spirulina-based production of heterologous proteins, improving volumetric productivity between 70% to 100% in multiple bioreactor setting configurations.
Conclusion An adaptive, machine learning-based approach to optimize heterologous protein production can improve outcomes based on complex, multivariate experiments, identifying beneficial variable combinations and adjustments that might not otherwise be discoverable within high-dimensional data.
Competing Interest Statement
J. R. is a founder and current employee of Lumen Bioscience, Inc. (Lumen) and owns stock/stock options of Lumen. C.G., J.D., J.M., and D.D. are employees of Lumen; all current and former employees own stock/stock options of Lumen. D. C. was an employee of Lumen at the time of data generation. Lumen has issued patents (U.S. 10,131,870) and a pending patent application (International Application No. PCT/US2020/040794) relating to certain research described in this article. D.B., E.B, L.C., M.B., and M.F. performed research as part of their employment at Google LLC. Google is a technology company that sells machine learning services as part of its business.