Genome-scale metabolic modelling when changes in environmental conditions affect biomass composition.

Area of Science:

• Systems biology
• Metabolic engineering

Background:

• Genome-scale metabolic models (GEMs) predict organism capabilities but often assume fixed biomass objective functions (BOFs).
• Cellular biomass composition varies with environmental conditions, challenging the accuracy of static BOFs.
• Accurate determination of biomass composition across environments is experimentally difficult.

Purpose of the Study:

• To develop and evaluate computational frameworks for integrating variable biomass composition data into GEMs.
• To address the need for mathematical approaches that handle dynamic cellular biomass.
• To improve the fidelity of metabolic predictions by accounting for environmental influences on biomass.

Main Methods:

• Proposed two novel computational approaches: Biomass Trade-off Weighting (BTW) and Higher-dimensional-plane InterPolation (HIP).
• Assessed BTW and HIP using three hypothetical BOFs within the *Escherichia coli* iML1515 GEM.
• Evaluated model performance and phenotypic predictions, including acetate secretion and respiratory quotient.

Main Results:

• Both BTW and HIP significantly impact GEM performance and predicted phenotypes.
• BTW consistently yielded higher growth rates across environments compared to HIP.
• HIP generated BOFs more closely resembling a reference BOF, while BTW showed greater differences.

Conclusions:

• The developed BTW and HIP methods offer a conceptual advance for GEMs.
• These frameworks enable GEMs to better reflect the environment-dependent nature of cellular biomass.
• Future work can refine these methods for more accurate metabolic modeling and predictions.