Non-Invasive Estimation of Acetates Using Off-Gas Information from Fed-Batch E. Coli Bioprocess

Pharmaceutical industries widely use Escherichia coli cell strain to synthesize various target products. The main goal is to reach the highest possible product yield. However, throughout the cell growth stage, the formation of by-products is inevitable. Metabolic compounds such as acetates cause inhibition, particularly in later bioprocess stages. The acetate accumulation model is necessary for planning bioprocesses to maximize cell biomass growth. This work depended on a black-box model. The decision tree method was in possession to replicate the approach. Specific biomass growth [1] at induction, broth weight, oxygen uptake rate [2], consumed substrate weight served for model training. Broth and consumed substrate weight had additional aging-related information [3] incorporated as separate inputs to introduce the cumulative regularization. Growth-limiting and non-growth-limiting dose feeding, bioprocess data used for black-box model training and validation. 80% of data served for model training. 20% functioned for model validation and 20% for both training validation. Mean absolute error (MAE) and residual sum of squares (RSS) were the criteria for validation. With used inputs and decision tree parameters, the best MAE was lower than 0.2, and the lowest RSS was lower than 2.

Funding: This project received funding from the European Regional Development Fund (project no. 01.2.2-LMT-K-718-03-0039) under a grant agreement with the Research Council of Lithuania (LMTLT).

References

1. Urniezius and A. Survyla, “Identification of Functional Bioprocess Model for Recombinant E. Coli Cultivation Process,” Entropy, vol. 21, no. 12, p. 1221, Dec. 2019, doi: 10.3390/e21121221.
2. Survyla, D. Levisauskas, R. Urniezius, and R. Simutis, “An oxygen-uptake-rate-based estimator of the specific growth rate in Escherichia coli BL21 strains cultivation processes,” Computational and Structural Biotechnology Journal, vol. 19, pp. 5856–5863, 2021, DOI: 10.1016/j.csbj.2021.10.015.
3. Urniezius, B. Kemesis, and R. Simutis, “Bridging Offline Functional Model Carrying Aging-Specific Growth Rate Information and Recombinant Protein Expression: Entropic Extension of Akaike Information Criterion,” Entropy, vol. 23, no. 8, p. 1057, Aug. 2021, doi: 10.3390/e23081057.