Astaxanthin biosynthesis: A two-step optimization approach and model construction with Response Surface Methodology and Artificial Neural Network

The first part of this research is investigating and comparing yield of a synthetic medium submerged three sugars (glucose, fructose and sucrose) at four different concentrations and solid fermentation systems with wheat bran and lentil waste for biosynthesis of astaxanthin (ASX) pigment by Xanthophyllomyces dendrorhous ATCC 24202 and Sporidiobolus salmonicolor ATCC 24259 microorganisms. The second part is modeling and optimizing the most efficient biosynthesis depending on waste, yeast and production variables consisted of moisture content, pH and temperature using a design matrix. The yields produced by X. dendrorhous were 51.88 µg of ASX/g glucose for the submerged medium with the least glucose, and 210.49 µg of ASX/g glucose for the wheat bran fermentation system. It was understood that the yield values of the submerged systems were lower and there was no requirement for the addition of any supplement to the waste systems. It was found that R 2 =0.9869 was the highest value with the maximum predicted ASX amount of 109.23 µg of ASX/g wheat bran with X. dendrorhous using Artificial Neural Network modeling and the moisture content was the most significant production parameter.