LK-DFBA: a linear programming-based modeling strategy for capturing dynamics and metabolite-dependent regulation in...

Area of Science:

• Systems biology
• Metabolic engineering
• Computational modeling

Background:

• Metabolomics provides valuable data for metabolic engineering.
• Traditional Flux Balance Analysis (FBA) models have limitations in integrating metabolomics data due to simplifying assumptions.
• There is a need for computational tools that can incorporate metabolite levels and regulation into strain design.

Purpose of the Study:

• To develop a novel metabolic modeling strategy, Linear Kinetics-Dynamic Flux Balance Analysis (LK-DFBA), that integrates metabolomics data.
• To retain the computational advantages of FBA while relaxing its assumptions to include metabolite concentrations and regulation.
• To improve the accuracy of metabolic engineering strain design tools.

Main Methods:

• Designed and implemented LK-DFBA, a modeling strategy based on Dynamic FBA (DFBA).
• Added strictly linear constraints for metabolism dynamics and regulation.
• Evaluated LK-DFBA using simulated noisy data from small and large in silico models, including E. coli central carbon metabolism.

Main Results:

• LK-DFBA reproduced metabolite concentration dynamics more effectively than ordinary differential equation models under realistic conditions.
• LK-DFBA showed smaller performance degradation when metabolite time course data were missing compared to other methods.
• Qualitatively reasonable results were obtained for the E. coli model, with explored parameterization structures offering trade-offs between computation time and accuracy.

Conclusions:

• LK-DFBA enables the calculation of metabolite concentrations and considers metabolite-dependent regulation.
• The framework retains computational advantages of FBA, offering a proof-of-principle for new metabolic modeling.
• LK-DFBA has the potential for creating genome-scale dynamic models and enhancing strain engineering tools.