Predictive modeling of Persian walnut (Juglans regia L.) in vitro proliferation media using machine learning approaches: a comparative study of ANN, KNN and GEP models.

BACKGROUND: Optimizing plant tissue culture media is a complicated process, which is easily influenced by genotype, mineral nutrients, plant growth regulators (PGRs), vitamins and other factors, leading to undesirable and inefficient medium composition. Facing incidence of different physiological disorders such as callusing, shoot tip necrosis (STN) and vitrification (Vit) in walnut proliferation, it is necessary to develop prediction models for identifying the impact of different factors involving in this process. In the present study, three machine learning (ML) approaches including multi-layer perceptron neural network (MLPNN), k-nearest neighbors (KNN) and gene expression programming (GEP) were implemented and compared to multiple linear regression (MLR) to develop models for prediction of in vitro proliferation of Persian walnut (Juglans regia L.). The accuracy of developed models was evaluated using coefficient of determination (R2), root mean square error (RMSE) and mean absolute error (MAE). With the aim of optimizing the selected prediction models, multi-objective evolutionary optimization algorithm using particle swarm optimization (PSO) technique was applied. RESULTS: Our results indicated that all three ML techniques had higher accuracy of prediction than MLR, for example, calculated R2 of MLPNN, KNN and GEP vs. MLR was 0.695, 0.672 and 0.802 vs. 0.412 in Chandler and 0.358, 0.377 and 0.428 vs. 0.178 in Rayen, respectively. The GEP models were further selected to be optimized using PSO. The comparison of modeling procedures provides a new insight into in vitro culture medium composition prediction models. Based on the results, hybrid GEP-PSO technique displays good performance for modeling walnut tissue culture media, while MLPNN and KNN have also shown strong estimation capability. CONCLUSION: Here, besides MLPNN and GEP, KNN also is introduced, for the first time, as a simple technique with high accuracy to be used for developing prediction models in optimizing plant tissue culture media composition studies. Therefore, selection of the modeling technique to study depends on the researcher's desire regarding the simplicity of the procedure, obtaining clear results as entire formula and/or less time to analyze.

High throughput mathematical modeling and multi-objective evolutionary algorithms for plant tissue culture media formulation: Case study of pear rootstocks.

Simplified prediction of the interactions of plant tissue culture media components is of critical importance to efficient development and optimization of new media. We applied two algorithms, gene expression programming (GEP) and M5' model tree, to predict the effects of media components on in vitro proliferation rate (PR), shoot length (SL), shoot tip necrosis (STN), vitrification (Vitri) and quality index (QI) in pear rootstocks (Pyrodwarf and OHF 69). In order to optimize the selected prediction models, as well as achieving a precise multi-optimization method, multi-objective evolutionary optimization algorithms using genetic algorithm (GA) and particle swarm optimization (PSO) techniques were compared to the mono-objective GA optimization technique. A Gamma test (GT) was used to find the most important determinant input for optimizing each output factor. GEP had a higher prediction accuracy than M5' model tree. GT results showed that BA (Γ = 4.0178), Mesos (Γ = 0.5482), Mesos (Γ = 184.0100), Micros (Γ = 136.6100) and Mesos (Γ = 1.1146), for PR, SL, STN, Vitri and QI respectively, were the most important factors in culturing OHF 69, while for Pyrodwarf culture, BA (Γ = 10.2920), Micros (Γ = 0.7874), NH4NO3 (Γ = 166.410), KNO3 (Γ = 168.4400), and Mesos (Γ = 1.4860) were the most important influences on PR, SL, STN, Vitri and QI respectively. The PSO optimized GEP models produced the best outputs for both rootstocks.

Combining gene expression programming and genetic algorithm as a powerful hybrid modeling approach for pear rootstocks tissue culture media formulation.

BACKGROUND: Predicting impact of plant tissue culture media components on explant proliferation is important especially in commercial scale for optimizing efficient culture media. Previous studies have focused on predicting the impact of media components on explant growth via conventional multi-layer perceptron neural networks (MLPNN) and Multiple Linear Regression (MLR) methods. So, there is an opportunity to find more efficient algorithms such as Radial Basis Function Neural Network (RBFNN) and Gene Expression Programming (GEP). Here, a novel algorithm, i.e. GEP which has not been previously applied in plant tissue culture researches was compared to RBFNN and MLR for the first time. Pear rootstocks (Pyrodwarf and OHF) were used as case studies on predicting the effect of minerals and some hormones in the culture medium on proliferation indices. RESULTS: Generally, RBFNN and GEP showed extremely higher performance accuracy than the MLR. Moreover, GEP models as the most accurate models were optimized using genetic algorithm (GA). The improvement was mainly due to the RBFNN and GEP strong estimation capability and their superior tolerance to experimental noises or improbability. CONCLUSIONS: GEP as the most robust and accurate prospecting procedure to achieve the highest proliferation quality and quantity has also the benefit of being easy to use.

Modeling and Optimizing a New Culture Medium for In Vitro Rooting of G×N15 Prunus Rootstock using Artificial Neural Network-Genetic Algorithm.

The main aim of the present investigation is modeling and optimization of a new culture medium for in vitro rooting of G×N15 rootstock using an artificial neural network-genetic algorithm (ANN-GA). Six experiments for assessing different media culture, various concentrations of Indole - 3- butyric acid, different concentrations of Thiamine and Fe-EDDHA were designed. The effects of five ionic macronutrients (NH4+, NO3-, Ca2+, K+ and Cl-) on five growth parameters [root number (RN), root length (RL), root percentage (R%), fresh (FW) and dry weight (DW)] were evaluated using the ANN-GA method. The R2 correlation values of 0.88, 0.88, 0.98, 0.94 and 0.87 between observed and predicted values were acquired for all five growth parameters, respectively. The ANN-GA results indicated that among the input variables, K+ (7.6) and NH4+ (4.4), K+ (7.7) and Ca2+ (2.8), K+ (36.7) and NH4+ (4.3), K+ (14.7) and NH4+ (4.4) and K+ (7.6) and NH4+ (4.3) had the highest values of variable sensitivity ratio (VSR) in the data set, for RN, RL, R%, FW and DW, respectively. ANN-GA optimized LS medium for G×N15 rooting contained optimized amounts of 1 mg L-1 IBA, 100, 150, or 200 mg L-1 Fe-EDDHA and 1.6 mg L-1 Thiamine. The efficiency of the optimized culture media was compared to other standard media for Prunus rooting and the results indicated that the optimized medium is more efficient than the others.

Development of an Artificial Neural Network as a Tool for Predicting the Targeted Phenolic Profile of Grapevine (Vitis vinifera) Foliar Wastes.

High performance liquid chromatography data related to the concentrations of 12 phenolic compounds in vegetative parts, measured at four sampling times were processed for developing prediction models, based on the cultivar, grapevine organ, growth stage, total flavonoid content (TFC), total reducing capacity (TRC), and total antioxidant activity (TAA). 12 Artificial neural network (ANN) models were developed with 79 input variables and different number of neurons in the hidden layer, for the prediction of 12 phenolics. The results confirmed that the developed ANN-models (R2 = 0.90 - 0.97) outperform the stepwise regression models (R2 = 0.05 - 0.78). Moreover, the sensitivity of the model outputs against each input variable was computed by using ANN and it was revealed that the key determinant of phenolic concentration was the source organ of the grapevine. The ANN prediction technique represents a promising approach to predict targeted phenolic levels in vegetative parts of the grapevine.

Mathematical Modeling and Optimizing of in Vitro Hormonal Combination for G × N15 Vegetative Rootstock Proliferation Using Artificial Neural Network-Genetic Algorithm (ANN-GA).

The efficiency of a hybrid systems method which combined artificial neural networks (ANNs) as a modeling tool and genetic algorithms (GAs) as an optimizing method for input variables used in ANN modeling was assessed. Hence, as a new technique, it was applied for the prediction and optimization of the plant hormones concentrations and combinations for in vitro proliferation of Garnem (G × N15) rootstock as a case study. Optimizing hormones combination was surveyed by modeling the effects of various concentrations of cytokinin-auxin, i.e., BAP, KIN, TDZ, IBA, and NAA combinations (inputs) on four growth parameters (outputs), i.e., micro-shoots number per explant, length of micro-shoots, developed callus weight (CW) and the quality index (QI) of plantlets. Calculation of statistical values such as R2 (coefficient of determination) related to the accuracy of ANN-GA models showed a considerably higher prediction accuracy for ANN models, i.e., micro-shoots number: R2 = 0.81, length of micro-shoots: R2 = 0.87, CW: R2 = 0.88, QI: R2 = 0.87. According to the results, among the input variables, BAP (19.3), KIN (9.64), and IBA (2.63) showed the highest values of variable sensitivity ratio for proliferation rate. The GA showed that media containing 1.02 mg/l BAP in combination with 0.098 mg/l IBA could lead to the optimal proliferation rate (10.53) for G × N15 rootstock. Another objective of the present study was to compare the performance of predicted and optimized cytokinin-auxin combination with the best optimized obtained concentrations of our other experiments. Considering three growth parameters (length of micro-shoots, micro-shoots number, and proliferation rate), the last treatment was found to be superior to the rest of treatments for G × N15 rootstock in vitro multiplication. Very little difference between the ANN predicted and experimental data confirmed high capability of ANN-GA method in predicting new optimized protocols for plant in vitro propagation.

Somaclonal variations and their applications in horticultural crops improvement.

The advancements made in tissue culture techniques has made it possible to regenerate various horticultural species in vitro as micropropagation protocols for commercial scale multiplication are available for a wide range of crops. Clonal propagation and preservation of elite genotypes, selected for their superior characteristics, require high degree of genetic uniformity amongst the regenerated plants. However, plant tissue culture may generate genetic variability, i.e., somaclonal variations as a result of gene mutation or changes in epigenetic marks. The occurrence of subtle somaclonal variation is a drawback for both in vitro cloning as well as germplasm preservation. Therefore, it is of immense significance to assure the genetic uniformity of in vitro raised plants at an early stage. Several strategies have been followed to ascertain the genetic fidelity of the in vitro raised progenies comprising morpho-physiological, biochemical, cytological and DNA-based molecular markers approaches. Somaclonal variation can pose a serious problem in any micropropagation program, where it is highly desirable to produce true-to-type plant material. On the other hand, somaclonal variation has provided a new and alternative tool to the breeders for obtaining genetic variability relatively rapidly and without sophisticated technology in horticultural crops, which are either difficult to breed or have narrow genetic base. In the present paper, sources of variations induced during tissue culture cycle and strategies to ascertain and confirm genetic fidelity in a variety of in vitro raised plantlets and potential application of variants in horticultural crop improvement are reviewed.