Optimization of actinomycin V production by Streptomyces triostinicus using artificial neural network and genetic...

Area of Science:

• Biotechnology
• Microbial Production
• Antibiotic Discovery

Background:

• Streptomyces triostinicus is a newly isolated strain with potential for antibiotic production.
• Actinomycin V is a valuable antibiotic class, but its production optimization is crucial.
• Previous studies have not reported actinomycin V production from this specific strain.

Purpose of the Study:

• To optimize medium components for enhanced actinomycin V production from Streptomyces triostinicus.
• To apply artificial neural network (ANN) and genetic algorithm (GA) for optimizing antibiotic yield.
• To compare the efficacy of ANN/GA with response surface methodology (RSM).

Main Methods:

• Central composite design (CCD) was used for experimental data generation.
• An artificial neural network (ANN) model was built using experimental data.
• A genetic algorithm (GA) was employed to optimize ANN model inputs for maximum yield.
• Five medium components (MgSO(4), NaCl, glucose, soybean meal, CaCO(3)) were optimized.

Main Results:

• The ANN/GA model predicted optimal concentrations for five medium components.
• Maximum actinomycin V yield of 452.0 mg l(-1) was achieved at optimized concentrations.
• The ANN/GA method resulted in a 36.7% higher antibiotic yield compared to RSM.

Conclusions:

• ANN and GA are effective tools for optimizing fermentation medium components.
• The developed ANN/GA model significantly enhances actinomycin V production from Streptomyces triostinicus.
• This approach offers a superior alternative to RSM for microbial product yield optimization.