PYF: a multi-functional algorithm for predicting production and optimizing metabolic engineering strategy in...

Area of Science:

• Synthetic biology
• Metabolic engineering
• Computational biology

Background:

• Accurate simulation of microbial consortia is vital for optimizing metabolic engineering and increasing yields.
• Existing models often fail to capture the complex interplay between biosynthesis and self-growth, limiting predictive power.

Purpose of the Study:

• To introduce the Polymicrobial cell factory Yield Forecasting (PYF) algorithm for improved simulation of microbial consortia.
• To enhance the prediction accuracy of production in microbial cell factories by considering biosynthesis pathway expression.

Main Methods:

• Developed the PYF algorithm, integrating biosynthesis pathway expression degrees to model microbial interactions.
• Validated PYF using *Escherichia coli*-*E. coli* consortia across diverse conditions (mono-metabolite, dual-carbon, dual-metabolite exchange).
• Performed sensitivity analysis for metabolic engineering strategy optimization.

Main Results:

• PYF demonstrated high accuracy in predicting microbial production, with a mean relative error (MRE) of 0.106.
• Achieved an average determination coefficient of 0.883 and hypothesis testing parameter of 0.930.
• PYF reduced MRE by approximately 61.6% compared to existing algorithms and does not require enzyme catalytic data.

Conclusions:

• PYF offers a simplified and accurate approach for simulating polymicrobial systems, enhancing metabolic engineering.
• The algorithm effectively predicts production and optimizes strategies without complex network integration.
• PYF represents a novel method for advancing microbial consortium-based production.