Isotope-assisted metabolic flux analysis as an equality-constrained nonlinear program for improved scalability and...

Area of Science:

• Metabolic Engineering
• Systems Biology
• Computational Biology

Background:

• Metabolic flux analysis (MFA) estimates carbon flow in metabolic networks.
• Current methods face scalability challenges with increasing network complexity.
• Efficient computational methods are crucial for robust parameter estimation.

Purpose of the Study:

• To develop and validate an efficient computational framework for MFA.
• To enhance the scalability and robustness of MFA for large metabolic networks.
• To integrate isotopically-nonstationary MFA (inst-MFA) into a unified optimization approach.

Main Methods:

• Formulated MFA as an equality-constrained nonlinear program (NLP).
• Utilized an algebraic modeling language (AML) with advanced optimization solvers (e.g., CONOPT in GAMS).
• Employed collocation for isotopically-nonstationary MFA (inst-MFA) ODE systems, transcribing them into algebraic constraints.

Main Results:

• The NLP approach demonstrated superior scalability and robustness compared to traditional shooting methods.
• The framework effectively handles both steady-state and isotopically-nonstationary MFA (inst-MFA).
• Developed eiFlux software in Python and GAMS to implement the NLP approach.

Conclusions:

• The NLP formulation provides a powerful and efficient method for metabolic flux analysis.
• This approach is particularly advantageous for large-scale and complex metabolic models, including eukaryotic and co-culture systems.
• The eiFlux software facilitates the application of advanced MFA techniques to diverse biological systems.