A non-dominated sorting Differential Search Algorithm Flux Balance Analysis (ndsDSAFBA) for in silico multiobjective...

Area of Science:

• Metabolic Engineering
• Systems Biology
• Computational Biology

Background:

• Metabolic engineering aims to enhance cellular functions by altering metabolic networks.
• In silico reaction knockout simulations are used to predict genetic perturbation effects on metabolite production.
• Existing methods often prioritize growth coupling, but the ultimate goal is increased production, and they typically yield single solutions, neglecting cellular multi-objective behavior.

Purpose of the Study:

• To develop a novel multi-objective optimization method for metabolic engineering.
• To identify reaction knockouts that simultaneously maximize metabolite production and growth rates.
• To address the limitations of single-objective optimization in metabolic engineering.

Main Methods:

• Developed a new method termed ndsDSAFBA (non-dominated sorting Differential Search Algorithm and Flux Balance Analysis).
• Incorporated Pareto dominance concepts to handle competing objectives (production and growth rates).
• Validated the method using three genome-scale metabolic models.

Main Results:

• Obtained a set of non-dominated solutions, each representing a distinct mutant strain with trade-offs between production and growth.
• Demonstrated that ndsDSAFBA outperforms single-objective optimization (SOO) and other multi-objective optimization (MOO) methods.
• Achieved superior results in terms of both production rate and growth rate.

Conclusions:

• The ndsDSAFBA method effectively identifies reaction knockouts for enhanced cellular production and growth.
• Multi-objective optimization using Pareto dominance is crucial for realistic metabolic engineering strategies.
• This approach provides a more comprehensive set of solutions compared to traditional methods.