Machine learning-driven optimization of metabolic balance for β-carotene production

Area of Science:

• Metabolic Engineering
• Synthetic Biology
• Biotechnology

Background:

• Balancing metabolic pathways is crucial for efficient microbial production of valuable compounds.
• In Yarrowia lipolytica engineered for beta-carotene synthesis, lipid synthesis competes with carotenoid production for cellular resources.
• Precise regulation of resource allocation is necessary for optimal bioproduct yield.

Purpose of the Study:

• To establish a machine learning framework for understanding and optimizing beta-carotene production in Yarrowia lipolytica.
• To capture complex interactions among the mevalonate pathway, lipid synthesis, and beta-carotene synthesis modules.
• To guide iterative gene integration for enhanced beta-carotene titers.

Main Methods:

• Development of a machine learning framework to model metabolic interactions.
• Prediction of beta-carotene output based on gene combinations.
• Iterative gene integration guided by computational predictions.

Main Results:

• The machine learning framework successfully modeled interactions between key metabolic modules.
• Computational guidance led to iterative gene integration strategies.
• The best-performing strain (YLT226) exhibited a 7-fold increase in beta-carotene titer compared to the initial strain (YLT001).

Conclusions:

• The developed machine learning framework provides a powerful strategy for engineering metabolic flux distributions.
• This approach enables precise regulation of resource allocation for enhanced bioproduct synthesis.
• The study demonstrates a successful method for optimizing beta-carotene production in microbial platforms.