Employing active learning in the optimization of culture medium for mammalian cells

Abstract

Medium optimization is a crucial step during cell culture for biopharmaceutics and regenerative medicine; however, this step remains challenging, as both media and cells are highly complex systems. Here, we addressed this issue by employing active learning. Specifically, we introduced machine learning to cell culture experiments to optimize culture medium. The cell line HeLa-S3 and the gradient-boosting decision tree algorithm were used to find optimized media as pilot studies. To acquire the training data, cell culture was performed in a large variety of medium combinations. The cellular NAD(P)H abundance, represented as A450, was used to indicate the goodness of culture media. In active learning, regular and time-saving modes were developed using culture data at 168 h and 96 h, respectively. Both modes successfully fine-tuned 29 components to generate a medium for improved cell culture. Intriguingly, the two modes provided different predictions for the concentrations of vitamins and amino acids, and a significant decrease was commonly predicted for fetal bovine serum (FBS) compared to the commercial medium. In addition, active learning-assisted medium optimization significantly increased the cellular concentration of NAD(P)H, an active chemical with a constant abundance in living cells. Our study demonstrated the efficiency and practicality of active learning for medium optimization and provided valuable information for employing machine learning technology in cell biology experiments.

Active learning for medium optimization for selective bacterial culture

“…The present study first demonstrated ML-assisted medium specialization for differentiated bacterial growth. ML was remarkably significant in medium optimization for microbial and mammalian cells 21,26 , which could be widely applied to synthetic construction and production 43,44 . The present study provided alternative applications in medium development for selective culture.…”

Efforts to Minimise the Bacterial Genome as a Free-Living Growing System

“…It predicts the growth fitness upon novel medium compositions made artificially and selects those in which the predicted growth fitness is significantly improved. Machine learning prediction for better activity has successfully improved the cellular activity of mammalian cells [ 99 ], the translation activity of bacteria or cell-free systems [ 100 , 101 ], and the activity of bacterial secondary metabolism [ 102 ]. The successful applications strongly suggest that machine learning-mediated methodologies can benefit genome reduction or minimisation.…”

“…Repeated rounds of active learning improve the prediction accuracy of machine learning and the growth fitness in the finetuned medium. The availability and efficiency of active learning for medium optimization have been demonstrated in the case of the mammalian cell [ 99 ]. Active learning has also been utilized in drug discovery [ 105 ], structural biology [ 106 ], and translational activity in cell-free systems [ 101 ].…”

Optimisation of surfactin yield inBacillususing active learning and high-throughput mass spectrometry

“…Recently, synthetic biologists have begun to explore its application in various use cases; early examples include synthetic gene design [43], as well as automated optimisation of metabolic engineering tasks [44-47], media optimization in cell-free systems [48], an in silico design of genetic control circuits [49]. Various bespoke approaches have been developed for media optimisation in mammalian cell bioproduction [50-52], and recently several software packages have been developed for the optimisation of gene circuits and metabolic pathways [53]. Our work is a novel application of active learning in combination with metabolomic readouts and thus offers substantial opportunities for data- and cost-efficient optimisation of complex bioprocesses across scales.…”