Resolving cell composition through simple measurements, genome-scale modeling, and a genetic algorithm.

Cell composition data is essential for accurate genome-scale metabolic modeling. However, obtaining this data experimentally is labor-intensive and costly. Prior research has shown that flux balance analysis (FBA) depends heavily on precise biomass equations. No prior work had resolved how to derive these equations efficiently using minimal experimental input. This gap motivated the development of a computational method using a genetic algorithm. The method aims to reduce the need for extensive biochemical measurements. It leverages culture data like glucose uptake and growth rates. This approach could streamline metabolic modeling across diverse organisms.

The aim of this work is to develop a predictive method for cell composition that reduces reliance on costly experiments. The specific problem is the lack of efficient tools to derive biomass equations. The motivation comes from the need to improve FBA accuracy without extensive biochemical data. The method uses a genetic algorithm to optimize biomass equations. It integrates culture data with genome-scale models. This approach allows modeling under multiple growth conditions. The goal is to make metabolic flux modeling more accessible and efficient.

The method combines genome-scale metabolic models with a genetic algorithm for nonlinear optimization. It uses culture data such as glucose uptake and growth rates as inputs. A biomass equation is optimized to match known or inferred cell composition. The genetic algorithm iteratively adjusts biomass components to minimize prediction errors. The process is validated using experimentally determined (13)C flux data. The method is applied to Escherichia coli MG1655 under various growth environments. A second case study involves Clostridium acetobutylicum. The results are compared to known biomass equations from related organisms.

Optimization of the biomass equation improved flux predictions in Escherichia coli MG1655. The method produced biomass equations that matched (13)C flux data closely. Predicted fluxes through the TCA cycle aligned well with experimental measurements. The method also generated a biomass equation for Clostridium acetobutylicum. This equation was similar to one derived experimentally for Bacillus subtilis. The results suggest the method can work even with limited biochemical data. The genetic algorithm effectively adjusted biomass components. The approach reduced the need for extensive experimental measurements.

The method successfully predicted cell composition using a genetic algorithm and culture data. It improved flux predictions in Escherichia coli when compared to (13)C data. The approach also worked for Clostridium acetobutylicum with limited data. The results suggest the method can be applied to other organisms. The authors propose that this approach streamlines metabolic modeling. It reduces the need for costly biochemical experiments. The method aligns with existing FBA frameworks. The findings suggest potential for broader use in metabolic studies.