Skip to main content
Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Forecasting and optimizing Agrobacterium-mediated genetic transformation via ensemble model- fruit fly optimization algorithm: A data mining approach using chrysanthemum databases

  • Mohsen Hesami,

    Roles Conceptualization, Formal analysis, Writing – original draft, Writing – review & editing

    Affiliation Department of Plant Agriculture, Gosling Research Institute for Plant Preservation, University of Guelph, Guelph, ON, Canada

  • Milad Alizadeh,

    Roles Formal analysis, Methodology, Writing – original draft, Writing – review & editing

    Affiliation Department of Botany, University of British Columbia, Vancouver, BC, Canada

  • Roohangiz Naderi ,

    Roles Conceptualization, Data curation, Investigation, Methodology, Project administration, Supervision, Writing – review & editing

    Affiliation Department of Horticultural Science, Faculty of Agriculture, University of Tehran, Karaj, Iran

  • Masoud Tohidfar

    Roles Conceptualization, Data curation, Formal analysis, Software, Supervision, Validation, Visualization

    Affiliation Department of Plant Biotechnology, Faculty of Sciences & Biotechnology, Shahid Beheshti University, G.C., Tehran, Iran


Optimizing the gene transformation factors can be considered as the first and foremost step in successful genetic engineering and genome editing studies. However, it is usually difficult to achieve an optimized gene transformation protocol due to the cost and time-consuming as well as the complexity of this process. Therefore, it is necessary to use a novel computational approach such as machine learning models for analyzing gene transformation data. In the current study, three individual machine learning models including Multi-Layer Perceptron (MLP), Adaptive Neuro-Fuzzy Inference System (ANFIS), and Radial Basis Function (RBF) were developed for forecasting Agrobacterium-mediated gene transformation in chrysanthemum based on eleven input variables including Agrobacterium strain, optical density (OD), co-culture period (CCP), and different antibiotics including kanamycin (K), vancomycin (VA), cefotaxime (CF), hygromycin (H), carbenicillin (CA), geneticin (G), ticarcillin (TI), and paromomycin (P). Consequently, best-obtained results were used in the fusion process by bagging method. Results showed that ensemble model with the highest R2 (0.83) had superb performance in comparison with all other individual models (MLP:063, RBF:0.69, and ANFIS: 0.74) in the validation set. Also, ensemble model was linked to Fruit fly optimization algorithm (FOA) for optimizing gene transformation, and the results showed that the maximum gene transformation efficiency (37.54%) can be achieved from EHA105 strain with 0.9 OD600, for 3.8 days CCP, 46.43 mg/l P, 9.54 mg/l K, 18.62 mg/l H, and 4.79 mg/l G as selection antibiotics and 109.74 μg/ml VA, 287.63 μg/ml CF, 334.07 μg/ml CA and 87.36 μg/ml TI as antibiotics in the selection medium. Moreover, sensitivity analysis demonstrated that input variables have a different degree of importance in gene transformation system in the order of Agrobacterium strain > CCP > K > CF > VA > P > OD > CA > H > TI > G. Generally, the developed hybrid model in this study (ensemble model-FOA) can be employed as an accurate and reliable approach in future genetic engineering and genome editing studies.


Horticulture plants including fruits, vegetables, grapes, and ornamental plants are raw material and used by people for food, either as edible products or for culinary ingredients, for medicinal use or ornamental and aesthetic purposes. They are a genetically very diverse group and play a major role in modern society and the economy [14]. Chrysanthemum (Dendranthema × grandiflorum) can be categorized as the second most economically important ornamental species due to its color and morphological diversity [5]. Moreover, chrysanthemum has been used as a model plant for color modification [6]. Conventional propagation and breeding approaches are not able to meet the increasing demands of the market for this valuable ornamental plant. Therefore, novel biotechnological methods such as genetic manipulation and gene editing such as CRISPR/Cas9 can be employed in order to satisfy the demands of consumers. Optimizing the gene transformation protocol can be considered as the first and foremost step in successful genetic engineering and gene editing studies [6, 7]. Many factors such as in vitro regeneration parameters (temperature, type and age of explant, quality and intensity of light, type and concentration of plant growth regulators, medium compositions), bacterial optical cell density, antibiotic and chemical stimulants concentrations, and inoculation duration (immersion time), play an important role in the efficiency of gene transformation [5]. Establishing an optimized protocol for genetic Agrobacterium-mediated transformation can be considered as a highly complex system, and it is critical to comprehend the effect of different factors prompting the T-DNA delivery into various explants [5, 8]. Subsequently, further analyses are essential to check T-DNA integration and stability and to achieve the efficiency parameter of gene transformation [9]. However, it is usually difficult to achieve an optimized gene transformation protocol due to the cost and time-consuming as well as the complexity of this process. Therefore, gene transformation can be considered as a multi-variable and non-linear biological process. Hence, conventional linear computational methods such as simple regression are not appropriate for analyzing biological systems such as gene transformation. Machine learning algorithms as a non-linear approach can be considered as a suitable computational methodology for predicting and optimizing different complex biological systems. Several studies have proved the usefulness of ANN for modeling and predicting in vitro culture processes such as in vitro secondary metabolite production, shoot proliferation and somatic embryogenesis [1016]. Nowadays, the necessity of increased precision and accuracy of machine learning algorithms has encouraged researchers to develop applicable methods such as ensemble approaches. The key idea of ensemble is fusing or combining data derived from fused information in order to provide more precise estimations in comparing with using individual model [17]. Many researchers in several fields of study have used ensemble models [1820]. At more complex features such as gene transformation, ensemble methods could be used to integrate the advantages and strengths of individual models. Several studies have demonstrated that ensemble models can be more reliable and accurate to model complex systems [1720]. Therefore, ensemble model can be considered as a reliable tool to help the handling of complex systems and to data mining. Data mining can be defined as the process of discovering and understanding previously unknown relationships and dependencies in datasets. In fact, data mining can be applied to generate and model rules able to enhance knowledge or further insight from experimental data [21].

However, difficulty in achieving an optimized solution can be considered as one of the demerit points of most machine learning algorithms [2229]. To overcome this bottleneck, Zhang et al. [30] employed the genetic algorithm (GA) as one of the common optimization algorithms for optimizing relative humidity, light duration, agar concentration, and culture temperature in order to maximize indirect shoot organogenesis in Cucumis melo. In another study, Non-dominated Sorting Genetic Algorithm-II (NSGA-II) was employed to optimize different types and concentrations of disinfectants as well as immersion time for maximizing explant viability and minimizing in vitro contamination in chrysanthemum [10]. However, most studies have found the optimized solution by trials and error [14, 3136]. Fruit fly optimization algorithm (FOA) suggested by Pan [37] is a new evolutionary optimization and computation approach. This novel optimization algorithm has the merits of being simple to comprehend and to be written into linguistic terms which is not too complex compared with other optimization algorithms [38]. Therefore, this study has attempted to apply the FOA to find the optimal levels of different factors involved in gene transformation.

In the current study, data mining by using ensemble strategy was employed to assess the effect and importance of different factors in Agrobacterium-mediated genetic transformation.

Data dispersed into several single chrysanthemum databases was assembled in order to model them and obtain further insight into the effect of different factors involved in chrysanthemum gene transformation. Furthermore, FOA was linked to the ensemble model to find the optimal level of factors involved in chrysanthemum gene transformation. According to the best of our knowledge, this study is the first report of the application of ensemble model in the field of genetic engineering.


Evaluating and comparing different individual (MLP, RBF, and ANFIS) models and ensemble method

Three individual models including MLP, RBF, and ANFIS were applied for forecasting gene transformation efficiency in chrysanthemum based on eleven inputs including Agrobacterium strain, optical density (OD), co-culture period (CCP), and different antibiotics including kanamycin (K), vancomycin (VA), cefotaxime (CF), hygromycin (H), carbenicillin (CA), geneticin (G), ticarcillin (TI), and paromomycin (P). In order to improve forecasting results, the best estimations obtained by three individual models were fused through the bagging method.

The efficiency of the individual and ensemble models was determined based on the assessment of forecasted and observed data. All the R2 of testing, training, and validation datasets were over 63%, 69%, and 73% for MLP, RBF, and ANFIS models, respectively (Table 1). According to Table 1, the ensemble model had the better predictive ability on forecasting gene transformation efficiency (R2 > 0.86, 079, and 0.83 for training, testing and validation sets, respectively) compared with individual models. The good fit of the ensemble model can be traced by the correlation between observed and forecasted data for gene transformation efficiency (Fig 1). Also, RMSE and MBE, same as R2, in ensemble model were better than individual models (Table 1). Based on the performance criteria that was mentioned in Table 1, ensemble model was able to efficiently explain the performances of Agrobacterium-mediated gene transformation to different studied factors.

Fig 1. Scatter plot of model predicted vs. observed data of chrysanthemum gene transformation efficiency by ensemble model.

(A) Training set, (B) Testing set, and (C) Validation set.

Table 1. Performance criteria of individual and ensemble models for gene transformation efficiency of chrysanthemum in training, testing, and validation processes.

Optimizing gene transformation through FOA

The aim of the current study not only was to forecast the gene transformation but also was to find an optimized level of Agrobacterium strain, OD, CCP, and different antibiotics including K, VA, CF, H, CA, G, TI, and P for the maximum Agrobacterium-mediated gene transformation efficiency in chrysanthemum. FOA was linked to ensemble model for achieving the optimal level of factors involved in gene transformation. The result of the optimization process was summarized in Table 2. According to Table 2, the maximum gene transformation efficiency (37.54%) can be achieved from EHA105 strain with 0.9 OD600, for 3.8 days CCP, 46.43 mg/l P, 9.54 mg/l K, 18.62 mg/l H, and 4.79 mg/l G as selection antibiotics and 109.74 μg/ml VA, 287.63 μg/ml CF, 334.07 μg/ml CA and 87.36 μg/ml TI as antibiotics in the selection medium.

Table 2. The results of optimization process via FOA for gene transformation efficiency of chrysanthemum.

Sensitivity analysis of the models

Databases were also used to determine the overall VSR for identifying the comparative rank of inputs. The results of sensitivity analysis were presented in Table 3. Based on sensitivity analysis, Agrobacterium-mediated gene transformation was more sensitive to Agrobacterium strain, followed by CCP, K, CF, VA, P, OD, CA, H, TI, and G.

Table 3. The results of sensitivity analysis on the developed ensemble model to rank the importance of factors involved in Agrobacterium-mediated gene transformation of chrysanthemums using GUS gene.


The Agrobacterium-mediated gene transformation of the chrysanthemum was widely studied by discovering the susceptibility of different chrysanthemum cultivars to Agrobacterium tumefaciens [5, 9]. However, several studies have reported some obstacles to establish and develop chrysanthemum gene transformation system such as chimeric plant regeneration consisting of both non-transgenic and transgenic tissues [39, 40], low efficiency of gene transformation [4143], and transgene inactivation [44]. Due to these difficulties and also the complex nature of the gene transformation system, there is a dire need to employ new computational methods to optimize this system. AI models can be considered as a reliable strategy to develop and optimize gene transformation protocols. Although there are no reports to use AI models in genetic engineering and genome editing, several studies have previously proved the reliability and accuracy of AI methodology to predict and optimize different in vitro culture processes such as in vitro sterilization [45, 46], callogenesis [34, 47, 48], cell growth and protoplast culture [49, 50], somatic embryogenesis [34, 51, 52], shoot regeneration [12, 5355], androgenesis [33], hairy root culture [56, 57], and rhizogenesis [58]. In the current study, MLP, RBF, ANFIS, and ensemble models, for the first time, were used to develop a suitable model for chrysanthemum gene transformation and compare their prediction accuracy. According to our results, ensemble model had more accuracy than individual models for modeling and predicting the system. Although there is no report regarding the application of AI models in gene transformation studies, in line with our results, comparative studies in other fields revealed the better performance of ensemble models in comparison to individual models [1720]. On the other hand, one of the weaknesses of using AI models is that it is hard to obtain an optimized solution [10]. To tackle this problem, several studies [10, 11, 13, 45, 54] used GA and NSGA-II to optimize in vitro culture conditions. In the current study, FOA was linked to ensemble model for the optimization process. Based on our results, a hybrid ensemble model and FOA can be considered as an efficient computational methodology for predicting and optimizing Agrobacterium-mediated gene transformation.

Agrobacterium strains play a pivotal role in gene transformation [8]. Several studies showed that successfulness in chrysanthemum gene transformation directly depends on selecting a suitable strain [5, 9]. Ledger et al. [59] first tried to produce transgenic chrysanthemum through LBA4404, however, low transformation efficiency (1.7%) was observed. Just two years later, Renou et al. [42] reported that higher transformation frequency between 5% and 40% can be achieved by using EHA101. Further studies [60, 61] employed LBA4404 and EHA101 to compare the performance of these two strains on the chrysanthemum gene transformation. These studies [60, 61] showed that EHA101 caused to 8.8% gene transformation frequency whereas LBA4404 resulted in 5.2%. Afterward, the efficiency of EHA101 and EHA105 was studied and showed that EHA105 had better performance than EHA101 for chrysanthemum gene transformation [9]. In line with previous studies, our results elucidated that EHA105 is the best strain to obtain the maximum gene transformation frequency.

The selection marker is another factor that plays an important role in gene transformation systems [8]. Due to the fact that in the first study of chrysanthemum gene transformation [62], the neomycin phosphotransferase II (nptII) gene was applied as a selection marker, kanamycin has been the main selection antibiotic of transgenic chrysanthemums. However, a high level of kanamycin in the selection medium represses organogenesis due to the sensitivity of chrysanthemum to kanamycin [9]. Other antibiotics, such as geneticin, paromomycin, and hygromycin, have been successfully employed for the detection of transgenic cells of chrysanthemums [42, 61, 63]. Our results showed that the combination of 46.43 mg/l paromomycin, 9.54 mg/l kanamycin, 18.62 mg/l hygromycin, and 4.79 mg/l geneticin is the best antibiotics combination for the selection of transgenic tissues. In accordance with our results, Aida et al. [63] reported that paromomycin has less toxic to cells than other antibiotics such as kanamycin, and it can reduce the chance of non-transgenic chrysanthemums escapes. Also, our results showed that cefotaxime can be considered as the best antibiotic for the selection medium. Previous studies [42, 61, 63] have proved the usefulness of cefotaxime in the selection medium.

One of the most important factors in Agrobacterium-mediated gene transformation systems is the density of the Agrobacterium strain [5, 9]. Therefore, Optimizing the optimal bacterial inoculation density is very critical because, with higher OD levels, explants are completely colonized by Agrobacterium and, subsequently, bacteria elimination becomes more difficult [8]. Similar to the previous studies [60, 64, 65], our results indicated that transformation efficiency can be improved when an optical density (OD600) of 0.9 would be used. The co-cultivation period is expected to be another important factor in gene transformation and transgenic plant regeneration [8]. According to previous studies [9, 66, 67], the regeneration of chrysanthemum explants following cocultivation with A. tumefaciens was significantly decreased even when explants were cultured on optimized media. This negative impact was observed when a c-cultivation period of 8d was employed. According to our results, 3.8 days of co-cultivation is the best period for the gene transformation in the chrysanthemum. Similar results have been reported by Teixeira da Silva and Fukai [67] and Shinoyama et al. [9].


Recently, different individual AI models have been widely applied for modeling and predicting in vitro culture processes. In the current study, ensemble model for the first time was applied to model and predict gene transformation efficiency and to compare its accuracy with individual models. Our results showed that the ensemble model has better accuracy than MLP, RBF, and ANFIS for modeling and predicting complex systems such as Agrobacterium-mediated gene transformation. Also, FOA was able to accurately optimize the chrysanthemum's gene transformation. The results of the current study demonstrate that the developed hybrid model (Ensemble-FOA) can open a reliable and accurate window to a comprehensive study of the plant's biological processes.

Materials and methods

Case study and data collection

Several experimental databases were selected from previous studies where detailed descriptions of materials and methods are available [9, 3944, 59100]. Data supporting the effect of Agrobacterium strain, optical density (OD), co-culture period (CCP), and different antibiotics including kanamycin (K), vancomycin (VA), cefotaxime (CF), hygromycin (H), carbenicillin (CA), geneticin (G), ticarcillin (TI), and paromomycin (P) on gene transformation efficiency of chrysanthemum using GUS gene were summarized in Table 4.

Table 4. Studies on Agrobacterium-mediated gene transformation of chrysanthemums using GUS gene.

Modeling procedures

Three individual machine learning algorithms including Multi-Layer Perceptron (MLP), Adaptive Neuro-Fuzzy Inference System (ANFIS), and Radial Basis Function (RBF) were proposed as estimator tools for modeling and optimizing chrysanthemum gene transformation datasets. The input variables were Agrobacterium strain, OD, CCP, and different antibiotics including K, VA, CF, H, CA, G, TI, and P. Also, the efficiency of gene transformation was chosen as outputs. Databases were randomly divided into three datasets: training set (70% database), testing set (20% database), and validation set (10% database). The MLP as one of the well-know ANNs was employed according to Hesami et al. [45] procedure. Also RBF and ANFIS were employed according to Hesami et al. [10] and Hesami et al. [13] procedures.

Ensemble model

Ensemble is known as the process of combining and mixing data from various sources such as single outputs of several machine learning algorithms that the overall equation can be as follows; (1)

Where stands for target variable, x is a vector of independent estimators, ε stands for corresponding estimation error, and n is a number of observation data.

In order to develop ensemble models, Eq (1) can be introduced to the following form where several individual models are employed; (2)

Where m stands for the number of individual model and [] stands as matrix of estimations provided by each model.

Subsequently, the matrix of [] will be considered as input data infusion models.

Many methods have been recommended for fusing individual models, which reported that the most powerful and uncomplicated among different approaches is the bagging method for data fusing. Therefore, the best-resulted outputs achieved by three individual models were fused through the bagging method (Fig 2).

Fig 2. The schematic view of the proposed ensemble model.

Finally, the coefficient of determination (R2), Mean Bias Error (MBE), and Root Mean Square Error (RMSE) were employed to determine the predictive ability of the developed model.

Fruit fly optimization algorithm (FOA)

The FOA is a novel approach for selecting optimization based on the food-finding activities of the fruit fly (Fig 3). The fruit fly is a type of insect, which lives in the tropical and temperate regions and eats corrupt fruit. In the current study, the FOA was applied to find optimal levels of inputs for achieving the maximum gene transformation efficiency. The details of the FOA are presented as follows:

Step 1: Initialization parameters.

First, the maximum repeat number (maxgen), the initial fruit fly swarm location (X_axis,Y_axis), the population size (sizepop), and the random flight distance range (FR) should be considered. In this investigation, maxgen = 100, (X_axis, Y_axis) ⸦ [0,1], sizepop = 10, and FR ⸦ [–10,10] were considered.

Step 2: Evolution starting.

The generation = 0, and the random flight path and the route for food finding of a single fruit fly were considered.

Step 3: Preliminary computations.

The flight distance (Disti) of food finding of the fruit fly i were adjusted. Subsequently, the smell concentration decision value Si were determined. Si were entered into the GRNN model. Then, the fitness function value (also called the smell concentration Smelli) was assessed. The fitness function value was used as the root-mean-square error (RMSE) which calculates the deviation between the actual value and the forecasting value.

Step 4: Offspring generation.

The offspring generation is produced according to the following Equations: (3) (4) (5) (6) (7) (8) (9) (10) (11)

Then the offspring was linked to the ensemble model and the fitness function value again was determined. Also, generation = generation + 1 was considered.

Step 5: Circulation stops.

When the generation attains the maximum repeat number, the stop criterion would be satisfied, and the optimized parameter value of the ensemble model can be reached. Otherwise, the optimization process should go back to Step 2.

Fig 3. The schematic view of the fruit fly optimization algorithm (FOA).

Sensitivity analysis

Sensitivity analysis was conducted to identify the importance degree of input variables on the efficiency of gene transformation. The sensitivity of these parameters was measured by the criteria including variable sensitivity error (VSE) value displaying the performance (RMSE) of the ensemble model when that input variable is removed from the model. Variable sensitivity ratio (VSR) value was determined as ratio of VSE and ensemble model error (RMSE value) when all input variables are available. A higher important variable in the model was detected by higher VSR.

MATLAB (Matlab, 2010) software was employed to write codes and run the models.


  1. 1. Eyduran SP, Akin M, Ercisli S, Eyduran E, Maghradze D. Sugars, organic acids, and phenolic compounds of ancient grape cultivars (Vitis vinifera L.) from Igdir province of Eastern Turkey. Biological Research. 2015;48(1):2. doi: pmid:25654659
  2. 2. GÜNEY M, Kafkas S, Koc A, Aras S, KELEŞ H, Karci H. Characterization of quince (Cydonia oblonga Mill.) accessions by simple sequence repeat markers. Turkish Journal of Agriculture and Forestry. 2019;43(1):69–79. doi:
  3. 3. Marsic NK, Necemer M, Veberic R, Ulrih NP, Skrt M. Effect of cultivar and fertilization on garlic yield and allicin content in bulbs at harvest and during storage. Turkish Journal of Agriculture and Forestry. 2019;43(4):414–29. doi:
  4. 4. Gecer MK, Kan T, Gundogdu M, Ercisli S, Ilhan G, Sagbas HI. Physicochemical characteristics of wild and cultivated apricots (Prunus armeniaca L.) from Aras valley in Turkey. Genetic Resources and Crop Evolution. 2020;67(4):935–45. doi:
  5. 5. da Silva JAT. Chrysanthemum: advances in tissue culture, cryopreservation, postharvest technology, genetics and transgenic biotechnology. Biotechnology Advances. 2003;21(8):715–66. doi: pmid:14563477
  6. 6. Noda N, Yoshioka S, Kishimoto S, Nakayama M, Douzono M, Tanaka Y, et al. Generation of blue chrysanthemums by anthocyanin B-ring hydroxylation and glucosylation and its coloration mechanism. Sci Advanc. 2017;3(7):e1602785. doi: pmid:28782017
  7. 7. da Silva JAT, Kulus D. Chrysanthemum biotechnology: discoveries from the recent literature. Folia Horticulturae. 2014;26(2):67–77. doi:
  8. 8. Niazian M. Application of genetics and biotechnology for improving medicinal plants. Planta. 2019;249(4):953–73. doi: pmid:30715560
  9. 9. Shinoyama H, Aida R, Ichikawa H, Nomura Y, Mochizuki A. Genetic engineering of chrysanthemum (Chrysanthemum morifolium): current progress and perspectives. Plant Biotechnology. 2012;29:323–37. doi:
  10. 10. Hesami M, Naderi R, Tohidfar M. Modeling and Optimizing Medium Composition for Shoot Regeneration of Chrysanthemum via Radial Basis Function-Non-dominated Sorting Genetic Algorithm-II (RBF-NSGAII). Scientific Reports. 2019;9(1):1–11. doi: pmid:30626917
  11. 11. Jamshidi S, Yadollahi A, Arab MM, Soltani M, Eftekhari M, Sabzalipoor H, et al. Combining gene expression programming and genetic algorithm as a powerful hybrid modeling approach for pear rootstocks tissue culture media formulation. Plant Methods. 2019;15(1):136. doi: pmid:31832078
  12. 12. Barone JO. Use of multiple regression analysis and artificial neural networks to model the effect of nitrogen in the organogenesis of Pinus taeda L. Plant Cell, Tissue and Organ Culture. 2019;137(3):455–64. doi:
  13. 13. Hesami M, Naderi R, Tohidfar M, Yoosefzadeh-Najafabadi M. Application of adaptive neuro-fuzzy inference system-non-dominated sorting genetic Algorithm-II (ANFIS-NSGAII) for modeling and optimizing somatic embryogenesis of Chrysanthemum. Frontiers in Plant Science. 2019;10:869. doi: pmid:31333705
  14. 14. Nezami-Alanagh E, Garoosi G- A, Maleki S, Landín M, Gallego PP. Predicting optimal in vitro culture medium for Pistacia vera micropropagation using neural networks models. Plant Cell, Tissue and Organ Culture. 2017;129(1):19–33. doi:
  15. 15. Khvatkov P, Chernobrovkina M, Okuneva A, Dolgov S. Creation of culture media for efficient duckweeds micropropagation (Wolffia arrhiza and Lemna minor) using artificial mathematical optimization models. Plant Cell, Tissue and Organ Culture. 2019;136(1):85–100. doi: -.
  16. 16. Akin M, Eyduran SP, Eyduran E, Reed BM. Analysis of macro nutrient related growth responses using multivariate adaptive regression splines. Plant Cell, Tissue and Organ Culture. 2020;140:661–70. doi:
  17. 17. Alizadeh MR, Nikoo MR. A fusion-based methodology for meteorological drought estimation using remote sensing data. Remote Sensing of Environment. 2018;211:229–47. doi:
  18. 18. Hararuk O, Zwart JA, Jones SE, Prairie Y, Solomon CT. Model‐data fusion to test hypothesized drivers of lake carbon cycling reveals importance of physical controls. Journal of Geophysical Research: Biogeosciences. 2018;123(3):1130–42. doi:
  19. 19. Aiello G, Giovino I, Vallone M, Catania P, Argento A. A decision support system based on multisensor data fusion for sustainable greenhouse management. Journal of Cleaner Production. 2018;172:4057–65. doi:
  20. 20. Wu X- M, Zhang Q- Z, Wang Y- Z. Traceability of wild Paris polyphylla Smith var. yunnanensis based on data fusion strategy of FT-MIR and UV–Vis combined with SVM and random forest. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy. 2018;205:479–88. doi: pmid:30059874
  21. 21. Gago J, Pérez-Tornero O, Landín M, Burgos L, Gallego PP. Improving knowledge of plant tissue culture and media formulation by neurofuzzy logic: a practical case of data mining using apricot databases. Journal of Plant Physiology. 2011;168(15):1858–65. doi: pmid:21676490
  22. 22. Moravej M. Discussion of “Modified Firefly Algorithm for Solving Multireservoir Operation in Continuous and Discrete Domains” by Irene Garousi-Nejad, Omid Bozorg-Haddad, and Hugo A. Loáiciga. J Water Resour Plan Manag. 2017;143(10):07017004. doi: /(ASCE)WR.1943-5452.0000836.
  23. 23. Moravej M, Amani P, Hosseini-Moghari S- M. Groundwater level simulation and forecasting using interior search algorithm-least square support vector regression (ISA-LSSVR). Groundw Sustain. 2020:100447. doi:
  24. 24. Araghinejad S, Fayaz N, Hosseini-Moghari S- M. Development of a Hybrid Data Driven Model for Hydrological Estimation. Water Resour Manag. 2018;32(11):3737–50. doi:
  25. 25. Fayaz N, Condon LE, Chandler DG. Evaluating the Sensitivity of Projected Reservoir Reliability to the Choice of Climate Projection: A Case Study of Bull Run Watershed, Portland, Oregon. Water Resour Manag. 2020;34(6):1991–2009. doi:
  26. 26. Dezfooli D, Abdollahi B, Hosseini-Moghari S- M, Ebrahimi K. A comparison between high-resolution satellite precipitation estimates and gauge measured data: case study of Gorganrood basin, Iran. J WATER SUPPLY RES T. 2018;67(3):236–51. doi:
  27. 27. Soleimani S, Haddad OB, Moravej M. Modeling water quality parameters using data-driven methods. J Water, Soil. 2016;30(3):Pe743–Pe57.
  28. 28. Salehi M, Farhadi S, Moieni A, Safaie N, Ahmadi H. Mathematical Modeling of Growth and Paclitaxel Biosynthesis in Corylus avellana Cell Culture Responding to Fungal Elicitors using Multilayer Perceptron-Genetic Algorithm. Front Plant Sci. 2020;11:1148. doi: pmid:32849706
  29. 29. Hesami M, Condori-Apfata JA, Valencia MV, Mohammadi M. Application of Artificial Neural Network for Modeling and Studying In Vitro Genotype-Independent Shoot Regeneration in Wheat. App Sci. 2020;10:5370. doi:
  30. 30. Zhang Q, Deng D, Dai W, Li J, Jin X. Optimization of culture conditions for differentiation of melon based on artificial neural network and genetic algorithm. Sci Rep. 2020;10(1):3524. doi: pmid:32103071
  31. 31. Alanagh EN, Garoosi G-a, Haddad R, Maleki S, Landín M, Gallego PP. Design of tissue culture media for efficient Prunus rootstock micropropagation using artificial intelligence models. Plant Cell, Tissue and Organ Culture. 2014;117(3):349–59. doi:
  32. 32. Gago J, Martinez-Nunez L, Landin M, Flexas J, Gallego PP. Modeling the effects of light and sucrose on in vitro propagated plants: a multiscale system analysis using artificial intelligence technology. PloS One. 2014;9(1):e85989. doi: pmid:24465829
  33. 33. Niazian M, Shariatpanahi ME, Abdipour M, Oroojloo M. Modeling callus induction and regeneration in an anther culture of tomato (Lycopersicon esculentum L.) using image processing and artificial neural network method. Protoplasma. 2019;256(5):1317–32. doi: pmid:31055656
  34. 34. Niazian M, Sadat-Noori SA, Abdipour M, Tohidfar M, Mortazavian SMM. Image processing and artificial neural network-based models to measure and predict physical properties of embryogenic callus and number of somatic embryos in ajowan (Trachyspermum ammi (L.) Sprague). In Vitro Cellular and Developmental Biology-Plant. 2018;54(1):54–68. doi:
  35. 35. Nezami-Alanagh E, Garoosi G- A, Landín M, Gallego PP. Combining DOE with neurofuzzy logic for healthy mineral nutrition of pistachio rootstocks in vitro culture. Frontiers in Plant Science. 2018;9:1474. doi: pmid:30374362
  36. 36. Nezami-Alanagh E, Garoosi G- A, Landin M, Gallego PP. Computer-based tools provide new insight into the key factors that cause physiological disorders of pistachio rootstocks cultured in vitro. Scientific Reports. 2019;9(1):1–15. doi: pmid:30626917
  37. 37. Pan W- T. A new Fruit Fly Optimization Algorithm: Taking the financial distress model as an example. Knowledge-Based Systems. 2012;26:69–74. doi:
  38. 38. Li H-z, Guo S, Li C-j, Sun J-q. A hybrid annual power load forecasting model based on generalized regression neural network with fruit fly optimization algorithm. Knowledge-Based Systems. 2013;37:378–87. doi:
  39. 39. Pavingerová D, Dostál J, Bísková R, Benetka V. Somatic embryogenesis and Agrobacterium-mediated transformation of chrysanthemum. Plant Science. 1994;97(1):95–101. doi:
  40. 40. Benetka V, Pavingerová D. Phenotypic differences in transgenic plants of chrysanthemum. Plant Breeding. 1995;114(2):169–73. doi:
  41. 41. de Jong J, Mertens MMJ, Rademaker W. Stable expression of the GUS reporter gene in chrysanthemum depends on binary plasmid T-DNA. Plant Cell Reports. 1994;14(1):59–64. doi: pmid:24194229
  42. 42. Renou JP, Brochard P, Jalouzot R. Recovery of transgenic chrysanthemum (Dendranthema grandiflora Tzvelev) after hygromycin resistance selection. Plant Science. 1993;89(2):185–97. doi:
  43. 43. Urban LA, Sherman JM, Moyer JW, Daub ME. High frequency shoot regeneration and Agrobacterium-mediated transformation of chrysanthemum (Dendranthema grandiflora). Plant Science. 1994;98(1):69–79. doi:
  44. 44. Takatsu Y, Nishizawa Y, Hibi T, Akutsu K. Transgenic chrysanthemum (Dendranthema grandiflorum (Ramat.) Kitamura) expressing a rice chitinase gene shows enhanced resistance to gray mold (Botrytis cinerea). Scientia Horticulturae. 1999;82(1):113–23. doi:
  45. 45. Hesami M, Naderi R, Tohidfar M. Modeling and optimizing in vitro sterilization of chrysanthemum via multilayer perceptron-non-dominated sorting genetic algorithm-II (MLP-NSGAII). Frontiers in Plant Science. 2019;10:282. doi: pmid:30923529
  46. 46. Ivashchuk OA, Fedorova V, Shcherbinina NV, Maslova EV, Shamraeva E. Microclonal propagation of plant process modeling and optimization of its parameters based on neural network. Drug Invention Today. 2018;10(3):3170–5.
  47. 47. Mansouri A, Fadavi A, Mortazavian SMM. An artificial intelligence approach for modeling volume and fresh weight of callus–A case study of cumin (Cuminum cyminum L.). Journal of Theoretical Biology. 2016;397:199–205. doi: pmid:26987421
  48. 48. Munasinghe SP, Somaratne S, Weerakoon SR, Ranasinghe C. Prediction of chemical composition for callus production in Gyrinops walla Gaetner through machine learning. Information Processing in Agriculture. 2020;7(2):1–12. doi:
  49. 49. Albiol J, Campmajó C, Casas C, Poch M. Biomass estimation in plant cell cultures: a neural network approach. Biotechnology Progress. 1995;11(1):88–92. doi:
  50. 50. Shiotani S, Fukuda T, Arai F, Takeuchi N, Sasaki K, Kinosita T. Cell recognition by image processing: recognition of dead or living plant cells by neural network. JSME International Journal. 1994;37(1):202–8. doi:
  51. 51. Molto E, Harrell RC, editors. Neural network classification of sweet potato embryos. Optics in Agriculture and Forestry; 1993: International Society for Optics and Photonics.
  52. 52. Zhang C, Timmis R, Hu W- S. A neural network based pattern recognition system for somatic embryos of Douglas fir. Plant Cell, Tissue and Organ Culture. 1999;56(1):25–35. doi:
  53. 53. Arab MM, Yadollahi A, Shojaeiyan A, Ahmadi H. Artificial neural network genetic algorithm as powerful tool to predict and optimize in vitro proliferation mineral medium for G× N15 rootstock. Frontiers in plant science. 2016;7:e1526.
  54. 54. Jamshidi S, Yadollahi A, Ahmadi H, Arab M, Eftekhari MJFips. Predicting in vitro culture medium macro-nutrients composition for pear rootstocks using regression analysis and neural network models. Frontiers in Plant Science. 2016;7:274. doi: pmid:27066013
  55. 55. Gupta SD, Pattanayak A. Intelligent image analysis (IIA) using artificial neural network (ANN) for non-invasive estimation of chlorophyll content in micropropagated plants of potato. In Vitro Cellular and Developmental Biology-Plant. 2017;53(6):520–6. doi:
  56. 56. Mehrotra S, Prakash O, Khan F, Kukreja A. Efficiency of neural network-based combinatorial model predicting optimal culture conditions for maximum biomass yields in hairy root cultures. Plant Cell Reports. 2013;32(2):309–17. doi: pmid:23143691
  57. 57. Osama K, Somvanshi P, Pandey AK, Mishra BN. Modelling of nutrient mist reactor for hairy root growth using artificial neural network. European Journal of Scientific Research. 2013;97(4):516–26.
  58. 58. Arab MM, Yadollahi A, Eftekhari M, Ahmadi H, Akbari M, Khorami SS. Modeling and Optimizing a New Culture Medium for In Vitro Rooting of G× N15 Prunus Rootstock using Artificial Neural Network-Genetic Algorithm. Scientific reports. 2018;8(1):e9977. doi: pmid:29967468
  59. 59. Ledger SE, Deroles SC, Given NK. Regeneration and Agrobacterium-mediated transformation of chrysanthemum. Plant Cell Reports. 1991;10(4):195–9. doi: pmid:24221545
  60. 60. Shinoyama H, Komano M, Nomura Y, Kazuma T. Stable Agrobacterium-mediated transformation of chrysanthemum (Dendranthema x grandiflorum (Ramat.) Kitamura). Bulletin of the Fukui Agricultural Experiment Station. 1998;35:13–21.
  61. 61. Shinoyama H, Komano M, Nomura Y, Nagai T. Introduction of delta-endotoxin gene of Bacillus thuringiensis to chrysanthemum [Dendranthema× grandiflorum (Ramat.) Kitamura] for insect resistance. Breeding Science. 2002;52(1):43–50. doi:
  62. 62. Lemieux C, Firoozabady E, Robinson K, editors. Agrobacterium-mediated transformation of chrysanthemum. Integration of in vitro techniques in ornamental plant breeding Proceedings, symposium; 1990: EUCARPIA.
  63. 63. Aida R, Ohira K, Tanaka Y, Yoshida K, Kishimoto S, Shibata M, et al. Efficient transgene expression in chrysanthemum, Dendranthema grandiflorum (Ramat.) Kitamura, by using the promoter of a gene for chrysanthemum chlorophyll-a/b-binding protein. Breeding Science. 2004;54(1):51–8. doi:
  64. 64. Fukai S, de Jong J, Rademaker W. Efficient genetic transformation of chrysanthemum (Dendranthema grandiflorum (Ramat.) Kitamura) using stem segments. Japanese Journal of Breeding. 1995;45(2):179–84. doi:
  65. 65. Annadana S, Mlynárová L, Udayakumar M, de Jong J, Nap J- P. The potato Lhca3.St.1 promoter confers high and stable transgene expression in chrysanthemum, in contrast to CaMV-based promoters. Molecular Breeding. 2002;8(4):335–44. doi:
  66. 66. de Jong J, Rademaker W, van Wordragen MF. Restoring adventitious shoot formation on chrysanthemum leaf explants following cocultivation with Agrobacterium tumefaciens. Plant Cell, Tissue and Organ Culture. 1993;32(3):263–70. doi:
  67. 67. Teixeira da Silva J, Fukai S. Increasing transient and subsequent stable transgene expression in chrysanthemum (Dendranthema× grandiflora (Ramat.) Kitamura) following optimization of particle bombardment and Agroinfection parameters. Plant Biotechnology. 2002;19:229–40. doi:
  68. 68. Jd Jong, Van Wordragen M, Rademaker W, editors. Early transformation events in Dendranthema grandiflora. Integration of in vitro techniques in ornamental plant breeding Proceedings, symposium, 10–14 November 1990; 1990: EUCARPIA.
  69. 69. van Wordragen MF, de Jong J, Huitema HBM, Dons HJM. Genetic transformation of Chrysanthemum using wild type Agrobacterium strains; strain and cultivar specificity. Plant Cell Reports. 1991;9(9):505–8. doi: pmid:24213790
  70. 70. Aida R, Tabei Y, Hirai M, Shibata M. Agrobacterium-mediated transformation of chrysanthemum. Breeding Science. 1992;42(Suppl 2):270–1.
  71. 71. Van Wordragen MF, De Jong J, Schornagel MJ, Dons HJM. Rapid screening for host-bacterium interactions in Agrobacterium-mediated gene transfer to chrysanthemum, by using the GUS-intron gene. Plant Science. 1992;81(2):207–14. doi:
  72. 72. Courtney-Gutterson N, Firoozabady E, Lemieux C, Nicholas J, Morgan A, Robinson K, et al. Production of genetically engineered color-modified chrysanthemum plants carrying a homologous chalcone synthase gene and their field performance. Acta Horticulturae. 1993;336:57–62. doi:
  73. 73. Lowe JM, Davey MR, Power JB, Blundy KS. A study of some factors affecting Agrobacterium transformation and plant regeneration of Dendranthema grandiflora Tzvelev (syn.Chrysanthemum morifolium Ramat.). Plant Cell, Tissue and Organ Culture. 1993;33(2):171–80. doi:
  74. 74. van Wordragen MF, Honée G, Dons HJM. Insect-resistant chrysanthemum calluses by introduction of a Bacillus thuringiensis crystal protein gene. Transgenic Research. 1993;2(3):170–80. doi: pmid:8353535
  75. 75. Courtney-Gutterson N, Napoli C, Lemieux C, Morgan A, Firoozabady E, Robinson KEP. Modification of Flower Color in Florist's Chrysanthemum: Production of a White–Flowering Variety Through Molecular Genetics. Bio/Technology. 1994;12(3):268–71. doi: pmid:7764487
  76. 76. de Jong J, Rademaker W, Ohishi K. Agrobacterium-mediated transformation of chrysanthemum. Plant Tissue Culture and Biotechnology. 1995;1:38–42.
  77. 77. Dolgov S, Mityshkina T, Rukavtsova E, Buryanov Y. Production of transgenic plants of chrysanthemum morifolium ramat with gene of Bac. thuringiensis delta-endotoxin. Acta Horticulturae. 1995;441:21–8. doi:
  78. 78. Oka S, Muraoka O, Abe T, Nakajima S. Formation of leaf-like bodies and adventitious buds, and chimeric expression of introduced GUS gene in garland chrysanthemum tissue cultures. Journal of the Japanese Society for Horticultural Science. 1996;65(2):294–5.
  79. 79. Dolgov S, Mitiouchkina T, Skryabin K. Agrobacterial transformation of chrysanthemum. Acta Horticulturae. 1997;447:329–34. doi:
  80. 80. Boase MR, Bradley JM, Borst NK. Genetic transformation mediated byAgrobacterium tumefaciens of florists' chrysanthemum (Dendranthema x grandiflorum) cultivar ‘Peach Margaret’. In Vitro Cellular & Developmental Biology—Plant. 1998;34(1):46–51. doi:
  81. 81. Boase MR, Butler RC, Borst NK. Chrysanthemum cultivar–Agrobacterium interactions revealed by GUS expression time course experiments. Scientia Horticulturae. 1998;77(1):89–107. doi:
  82. 82. Fu R- Z, Liu M, Liang H- J, Zhang C- H, Xue H, Sun Y- R. Production of transgenic plants of chrysanthemum via Agrobacterium tumefaciens mediated method. Acta Phytophysiologica. 1998;24:72–6.
  83. 83. Kim J, Park Y, Jung S, Chung H, Shin Y, Sheop J. Transformation of chrysanthemum by Agrobacterium tumefaciens with three different types of vectors. Korean Journal of Horticultural Science & Technology. 1998;39:360–6.
  84. 84. Kim M, Kim J, Hee Y. Plant Regeneration and Flavonoid 3'5'-Hydroxylase gene Tranformation of Dendranthema zawadskii and Dendranthema indicum. Korean Journal of Horticultural Science & Technology. 1998;39:355–9.
  85. 85. John MS, James WM, Margaret ED. A Regeneration and Agrobacterium-mediated Transformation System for Genetically Diverse Chrysanthemum Cultivars. Journal of the American Society for Horticultural Science jashs. 1998;123(2):189–94. doi:
  86. 86. Shinoyama H, Nomura Y, Tuchiya T, Kazuma T. Direct embryoid formation and plant regeneration from leaves of chrysanthemum (Dendranthema grandiflora tzveiev.). Breeding Science. 1996;46(2):158–9.
  87. 87. Takatsu Y, Tomotsune H, Kasumi M, Sakuma F. Differences in adventitious shoot regeneration capacity among Japanese chrysanthemum [Dendranthema grandiflorum (Ramat.) Kitamura] cultivars and the improved protocol for Agrobacterium-mediated genetic transformation. Journal of the Japanese Society for Horticultural Science. 1998;67(6):958–64. doi:
  88. 88. Young KJ, Jung PS, Young UB, Ho PC, Soo CY, Sheop SJ. Transformation of chrysanthemum by Agrobacterium tumefaciens with three different types of vectors. Horticulture Environment and Biotechnology 1998;39(3):360–6.
  89. 89. Shao H, Li J, Zheng X, Chen S. Cloning of the LFY cDNA from Arabidopsis thaliana and its transformation to Chrysanthemum morifolium. Acta Botanica Sinica. 1999;41(3):268–71. PubMed PMID: pmid:325954.
  90. 90. Shirasawa N, Iwai T, Nakamura S, Honkura R. Transformation and transgene expression of chrysanthemum [Dendranthema grandiflorum (Ramat) Kitamura]. Bulletin of the Miyagi Prefectural Agricultural Research Center. 2000;67:15–20.
  91. 91. Tosca A, Delledonne M, Furini A, Belenghi B, Fogher C, Frangi P. Transformation of Korean chrysanthemum (Dendranthema zawadskii× D.× grandiflorum) and insertion of the maize autonomous element Ac using Agrobacterium tumefaciens. Journal of Genetics and Breeding 2000;54(1):19–24.
  92. 92. Zhi-Liang Z, Zhenbiao Y, Jyan-Chyan J, James DM. Modification of Plant Architecture in Chrysanthemum by Ectopic Expression of the Tobacco Phytochrome B1 Gene. Journal of the American Society for Horticultural Science 2001;126(1):19–26. doi:
  93. 93. Ishida I, Tukahara M, Yoshioka M, Ogawa T, Kakitani M, Toguri T. Production of anti-virus, viroid plants by genetic manipulations. Pest Management Science. 2002;58(11):1132–6. doi: pmid:12449532
  94. 94. Jeong JH, Chakrabarty D, Kim SJ, Paek KY. Transformation of chrysanthemum (Dendranthema grandiflorum Kitamura cv. Cheonsu) by constitutive expression of rice OsMADS1 gene. Horticulture Environment and Biotechnology. 2002;43(4):382–6.
  95. 95. Kudo S, Shibata N, Kanno Y, Suzuki M. Transformation of chrysanthemum (Dendranthema grandiflorum (Ramat.) Kitamura) via Agrobacterium tumefaciens. Acta Horticulturae. 2002;572:139–47. doi:
  96. 96. Toguri T, Ogawa T, Kakitani M, TUKAHARA M, YOSHIOKA M. Agrobacterium-mediated transformation of chrysanthemum (Dendranthema grandiflora) plants with a disease resistance gene (pac1). Plant biotechnology. 2003;20(2):121–7. doi:
  97. 97. Petty LM, Harberd NP, Carré IA, Thomas B, Jackson SD. Expression of the Arabidopsis gai gene under its own promoter causes a reduction in plant height in chrysanthemum by attenuation of the gibberellin response. Plant Science. 2003;164(2):175–82. doi:
  98. 98. Outchkourov NS, Peters J, de Jong J, Rademakers W, Jongsma MA. The promoter–terminator of chrysanthemum rbcS1 directs very high expression levels in plants. Planta. 2003;216(6):1003–12. doi: pmid:12687368
  99. 99. Aida R, Nagaya S, Yoshida K, Kishimoto S, Shibata M, Ohmiya A. Efficient transgene expression in chrysanthemum, Chrysanthemum morifolium Ramat., with the promoter of a gene for tobacco [Nicotiana tabacum] elongation factor 1 alpha protein. Japan Agricultural Research Quarterly. 2005;39(4):269–74. doi:
  100. 100. Aida R, Narumi T, Ohtsubo N, Yamaguchi H, Kato K, Shinmyo A, et al. Improved translation efficiency in chrysanthemum and torenia with a translational enhancer derived from the tobacco alcohol dehydrogenase gene. Plant Biotechnology. 2008;25(1):69–75. doi: </References>