Advanced modeling and optimizing for surface sterilization process of grape vine (Vitis vinifera) root stock 3309C through response surface, artificial neural network, and genetic algorithm techniques
Habtamu Dagne,
Venkatesa Prabhu S,
Hemalatha Palanivel,
Alazar Yeshitila,
Solomon Benor,
Solomon Abera,
Adugna Abdi
Affiliations
Habtamu Dagne
Department of Biotechnology, Centre of Bioprocess and Biotechnology, College of Biological and Chemical Engineering, Addis Ababa Science and Technology University, PO Box 16417, Addis Ababa, Ethiopia
Venkatesa Prabhu S
Department of Chemical Engineering, Centre of Excellence for Biotechnology and Bioprocess, College of Biological and Chemical Engineering, Addis Ababa Science and Technology University, PO Box 16417, Addis Ababa, Ethiopia
Hemalatha Palanivel
Department of Biotechnology, Centre of Bioprocess and Biotechnology, College of Biological and Chemical Engineering, Addis Ababa Science and Technology University, PO Box 16417, Addis Ababa, Ethiopia; Corresponding author.
Alazar Yeshitila
Department of Biotechnology, Centre of Bioprocess and Biotechnology, College of Biological and Chemical Engineering, Addis Ababa Science and Technology University, PO Box 16417, Addis Ababa, Ethiopia
Solomon Benor
Department of Biotechnology, Centre of Bioprocess and Biotechnology, College of Biological and Chemical Engineering, Addis Ababa Science and Technology University, PO Box 16417, Addis Ababa, Ethiopia; Department of Plant Biology and Biodiversity Management, College of Natural and Computational Sciences, Addis Ababa University, Ethiopia; Department of Industrial Engineering, Faculty of Engineering and the Built, Environment, Tshwane University of Technology, South Africa
Solomon Abera
Department of Biotechnology, Centre of Bioprocess and Biotechnology, College of Biological and Chemical Engineering, Addis Ababa Science and Technology University, PO Box 16417, Addis Ababa, Ethiopia
Adugna Abdi
Department of Biotechnology, Centre of Bioprocess and Biotechnology, College of Biological and Chemical Engineering, Addis Ababa Science and Technology University, PO Box 16417, Addis Ababa, Ethiopia
In vitro, sterilization is one of the key components for proceeding with plant tissue cultures. Since the effectiveness of sterilization has a direct impact on the culture's final outcomes, there is a crucial need for optimization of the sterilization process. However, compared with traditional optimizing methods, the use of computational approaches through artificial intelligence-based process modeling and optimization algorithms provides a precise optimal condition for in vitro culturing. This study aimed to optimise in vitro sterilization of grape rootstock 3309C using RSM, ANN, and genetic algorithm (GA) techniques. In this context, two output responses, namely, Clean Culture and Explant Viability, were optimised using the models developed by RSM and ANN, followed by a GA, to obtain a globally optimal solution. The most influential independent factors, such as HgCl2, NaOCl, AgNO3, and immersion time, were considered input variables. The significance of the developed models was investigated with statistical and non-statistical techniques and was optimised to determine the significance of selected inputs. The optimal clean culture of 91%, and the explant viability of 89% can be obtained from 1.62% NaOCl at a 13.96 min immersion time, according to MLP-NSGAII. Sensitivity analysis revealed that the clean culture and explant viability were less sensitive to AgNO3 and more sensitive to immersion time. Results showed that the differences between the GA predicted and validation data were significant after the performance validation of predicted and optimised sterilising agents with immersion time combinations were tested. In general, GA, a potent methodology, may open the door to the development of new computational methods in plant tissue culture.
