Leveraging knowledge engineering and machine learning for microbial bio-manufacturing

Abstract

Genome scale modeling (GSM) predicts the performance of microbial workhorses and helps identify beneficial gene targets. GSM integrated with intracellular flux dynamics, omics, and thermodynamics have shown remarkable progress in both elucidating complex cellular phenomena and computational strain design (CSD). Nonetheless, these models still show high uncertainty due to a poor understanding of innate pathway regulations, metabolic burdens, and other factors (such as stress tolerance and metabolite channeling). Besides, the engineered hosts may have genetic mutations or non-genetic variations in bioreactor conditions and thus CSD rarely foresees fermentation rate and titer. Metabolic models play important role in design-build-test-learn cycles for strain improvement, and machine learning (ML) may provide a viable complementary approach for driving strain design and deciphering cellular processes. In order to develop quality ML models, knowledge engineering leverages and standardizes the wealth of information in literature (e.g., genomic/phenomic data, synthetic biology strategies, and bioprocess variables). Data driven frameworks can offer new constraints for mechanistic models to describe cellular regulations, to design pathways, to search gene targets, and to estimate fermentation titer/rate/yield under specified growth conditions (e.g., mixing, nutrients, and O). This review highlights the scope of information collections, database constructions, and machine learning techniques (such as deep learning and transfer learning), which may facilitate "Learn and Design" for strain development.

Engineering Robust Production Microbes for Large-Scale Cultivation

“…Miniaturization and automation of strain design, strain construction, and high-throughput measurements provide the data and data-driven approaches to translate production performance from laboratory to pilot and commercial scales. Implementation of advanced statistical methods, such as machine learning, can expedite strain-engineering cycles [80,97,98] and are being adopted by the biotech industry focused on high-throughput strain engineering [99,100]. The state of computational approaches and modeling has not been a focus of this review but has been discussed recently in several comprehensive articles [8,101].…”

Towards a widespread adoption of metabolic modeling tools in biopharmaceutical industry: a process systems biology engineering perspective

“…The application of artificial intelligence and machine learning in bioprocess engineering and systems biology has seen a significant increase in the last years thanks to the research advances in these fields [41][42][43] and their successful applications for e.g. in tumor detection 44 .…”

Systems Metabolic Engineering Meets Machine Learning: A New Era for Data‐Driven Metabolic Engineering

“…For example, a supervised ML algorithm used to predict the selling price of used cars would utilize a feature vector that considers the car make, age, mileage, and original purchase price to produce the output of the expected selling price. When applying such an approach to biological systems, the feature vector could take into account information in the ‘omics datasets above, physiological conditions, bioreactor conditions, and growth conditions in an attempt to predict an output variable such as growth rate or metabolite overproduction. In this regard, the predictions made by these algorithms can be a binary classification, a multiclass classification, or a continuous regression.…”