A hybrid of response surface methodology and artificial neural network in optimization of culture conditions of mycelia growth of Antrodia cinnamomea

Antrodia cinnamomea (A. cinnamomea) faces the challenge of coping with commercial usage in formulating nutraceuticals and functional foods in Taiwan. This research aimed to increase the biomass production of mycelia during the cultivation of A. cinnamomea using a methodology that hybrids Response Surface Methodology (RSM) and Artificial Neural Network (ANN). RMS aimed to optimize the culture condition while ANN intended to identify the factors dominating biomass production. The Plackett-Burman design and 32 (2(7-2)) fractional factorial designs identified four key factors. A four-factor six-level central composite design was used to investigate the correlation between the biomass and the key factors. The yield of RSM was 200% higher than that of the control medium. The proposed methodology offers reliable production of the medicinal fungus under optimum conditions in laboratory culture and reduces the cost, time and effort made, compared to the slow-growing propagation in nature. ANN opens a new opportunity of biomass prediction in microbial cultivation. Moreover, we provide the potential of hybrid RSM-ANN methods when encountering multifarious tasks in the future with the hope of bringing forward a new generation of biomass production technologies.

Automated summary

This research successfully increased the biomass production of Antrodia cinnamomea using a hybrid methodology of response surface methodology (RSM) and artificial neural network (ANN). The proposed method offers reliable production in the laboratory, reducing costs and time compared to natural propagation. This has important implications for biomass production technologies.