Short CommunicationApplication of genetic algorithm in modelling and optimization of cellulase production
Introduction
Significant amounts of by-products derived from agro-processing industries go as waste. Improper disposition of such agro-industrial wastes contribute to environmental pollution and are home to several micro-organisms. These wastes could be recycled or further processed to extract or formulate value-added products. Glucoamylase is a potent starch degrading enzyme whose cheap production has been an area of research. Its production is generally carried out by fungal cultures, in solid-state fermentation as well as submerged fermentation (Pandey, 1995, Pandey, 2003). The potential of glucoamylase and related technologies inspired the utilization of other enzymes of industrial importance. In light of this revelation, the production of cellulase using food waste was identified as a potent research area. Pea hulls obtained from pea processing industries contains 63.2% cellulose, 8.2% hemi-cellulose (Sosulski and Wu, 1988) and has been described as a good carbon source for cellulase production by fungal cultures (El-Shishtawy et al., 2015). It also comprises 60% of peapods. Optimization of bioprocesses for enzymes production based on genetics has evolved in the recent decade (Binod et al., 2017, Madhavan et al., 2017). Although it lacks a mathematical foundation, it greatly enhances the prediction capability of computer models. Genetic Algorithm (GA) is an Evolutionary Algorithm (EA), which works on Darwin’s theory of natural selection. In GA, genes are expressed as a combination of binary digits ‘1’ and ‘0’. A chromosome comprising of these genes (e.g., 1001–1101 or 1110–1010, etc.) are called strings. GA mixes and matches the independent variables to produce superior ‘offsprings’ thereby increasing flexibility, efficiency and efficacy of the model to be optimized. It performs three operations: selection, crossover and mutation. Initially strings (obtained from variables, Xi) are selected randomly as per higher relative fitness (value of response, Y). They are then allowed to enter the mating zone where random crossover points are chosen with a certain probability of crossover (Pc). Crossover implies the interchange of bit values (genes) i.e., 1 to 0 and vice versa. This continues until a particular population size is reached. Finally, mutation occurs when random genes within the new population are changed (1 to 0 and vice versa) with a probability of mutation (Pm). The result is a population (new generation) with improved average fitness. Detailed information for the steps involved in GA has been reported by various researchers (Konak et al., 2006, Blaifi et al., 2016). The objective of the present work was to exploit the potential of GA for determining suitable optimized parameters for the production of cellulase. The study also compared the results of GA to that of a mathematical optimizer.
Section snippets
Sample preparation
Green pea pods were procured from the local market of Pantnagar, Uttarakhand (India). Separation of pea hulls from the pods was done manually. Pea hull slurry was prepared by grinding the separated hulls at 18,000 rpm for 3 min in a commercial blender (make: Sujata) and strained through a 2 mm pore size mesh.
Microorganism and inoculum
Trichoderma reesei QM9414 strain was procured from IMTECH, Chandigarh, India and was used for the production of cellulase. Spore suspension was prepared by incubating the cultures on Potato
Results and discussion
Second order polynomial model for the observed data was generated for coded values of independent and dependent variables (Table 1). Coded values for variables were calculated using Eq. (3) where, Av, Cv, H and L represents the actual, coded, highest and lowest values respectively. Coded form of model equation (Eq. (4)) was preferred over the true form [Eq. (5)] so that unbiased comparison of most affecting variables can be done.
Based on the obtained model, the cultivation time
Conclusion
It could be concluded from the results that the application of Genetic Algorithm generated process conditions which resulted in increased cellulase activity as compared with mathematical optimizer. The investigation also highlighted that incubation time had a significant effect on the cellulase activity and would be therefore an important parameter to be considered in processes involving enzyme extraction. Pea hull was observed to have potential for cellulase production.
Acknowledgement
The authors thank Dr. A.K. Tiwari for providing laboratory facility for this work. Also, the first author wholeheartedly accepts the marriage proposal (NCAA, Springer) of the third author.
References (25)
- et al.
Recent developments in l-glutaminase production and applications – an overview
Biores. Technol
(2017) - et al.
An enhanced dynamic model of battery using genetic algorithm suitable for photovoltaic applications
Appl. Energy
(2016) - et al.
Multi-objective optimization using genetic algorithm: a tutorial
Rel. Eng. Syst. Saf.
(2006) - et al.
Evaluation of culture conditions for cellulase production by two Trichoderma reesei mutants under solid state fermentation conditions
Biores. Technol.
(2007) - et al.
Strategies for design of improved biocatalysts for industrial applications
Biores. Technol.
(2017) Solid-state fermentation
Biochem. Eng. J.
(2003)- et al.
Production of bioethanol, methane and heat from sugarcane bagasse in a biorefinery concept
Biores. Technol.
(2011) - et al.
Chemical pretreatment methods for the production of cellulosic ethanol: technologies and innovations
Int. J. Chem. Eng.
(2013) Optimization of pretreatment of low lignin content biomass
Appl. Biochem. Biotechnol.
(2009)- et al.
Studies on the effect of pH and carbon sources on enzyme activities of some pectinolytic bacteria isolated from jute retting water
Turk. J. Biol.
(2011)
Saccharification and hydrolytic enzyme production of alkali pre-treated wheat bran by Trichoderma virens under solid state fermentation
BMC Biotechnol.
Measurement of cellulase activities
Pure Appl. Chem.
Cited by (28)
Multi-objective scheduling and optimization for smart energy systems with energy hubs and microgrids
2024, Engineering Science and Technology, an International JournalAnalysis and optimization of thermal power plants: Pinch and exergy methods
2023, Journal of Cleaner ProductionMicrobial enzyme bioprocesses in biobleaching of pulp and paper: technological updates
2023, Microbial Bioprocesses: Applications and PerspectivesImproved deinking and biobleaching efficiency of enzyme consortium from Thermomyces lanuginosus VAPS25 using genetic Algorithm-Artificial neural network based tools
2022, Bioresource TechnologyCitation Excerpt :These techniques evaluate the critical factors, their calculated response at the most favorable concentration, and the outcome of their relations on production. Many recent findings have signified ANN united with GA to be on par with response surface methodology (RSM) or even an enhanced substitute in predicting optimization factors for diverse bioprocesses (Sirohi et al., 2018; Kumar et al., 2017). The use of enzymes produced from microorganisms in the industry is challenging and requires specific unique characteristics (Sirohi et al., 2018).
Optimisation of a smart energy hub with integration of combined heat and power, demand side response and energy storage
2021, EnergyCitation Excerpt :The controlled variables include both real and integer variables; thus, the optimisation is a mixed-integer programming problem. A stochastic global search method, Genetic Algorithm (GA) [43], is selected to solve this problem. GA simulates the biological processes of reproduction and the natural selection to find the optimal solution, which is proved to have good convergence rate, low computational time and high robustness with satisfactory precision [44].