Search

Abstract: Glucosinolates, a class of specialized metabolites specific to the order Brassicales, have diverse bioactivities that are largely dependent on the structures of their side chains. Flavin‐containing monooxygenases (FMOs) encoded by the FMOGS‐OX genes have been found to catalyze side‐chain modifications during the synthesis of methionine‐derived aliphatic glucosinolates. Seven FMOGS‐OX genes have been identified in Arabidopsis Heynh., but the evolution of these genes in the Brassicaceae, a family including many economically important vegetables, is poorly understood. In this study, the phylogenetic and syntenic relationships of the FMOGS‐OX genes belonging to 12 sequenced Brassicaceae species were analyzed. Our results showed that the FMOGS‐OX genes included two tandem arrays, the FMOGS‐OX2‐4 group (group A) and the FMOGS‐OX5‐7 group (group B). The evolutionary histories of the FMOGS‐OX groups A and B were similar across the Brassicaceae, but two lineage‐specific evolutionary routes developed after these two separate species lineages diverged from Aethionema arabicum (L.) Andrz. ex DC. In the lineage I route, FMOGS‐OX gene copies tended to increase due to frequent tandem duplication events in most species and a whole genome triplication in Camelina sativa (L.) Crantz. In the lineage II route, gene copies decreased due to deletion events. Combining these results with those of previous studies, we speculated that the FMOGS‐OX genes were derived from an ancestral gene with a broad expression distribution and a broad range of substrates, which then underwent subfunctionalization to generate progeny limited in either spatial expression or substrate structure. Furthermore, the absence of FMOGS‐OX5 substrates in some FMOGS‐OX5‐containing species may suggest neofunctionalization of these genes. [ABSTRACT FROM AUTHOR]

Hyperspectral imaging combined with feature extraction methods were applied to determine soluble sugar content (SSC) in mature and scatheless strawberry. Hyperspectral images of 154 strawberries covering the spectral range of 874-1,734 nm were captured and the spectral data were extracted from the hyperspectral images, and the spectra of 941~1,612 nm were preprocessed by moving average (MA). Nineteen samples were defined as outliers by the residual method, and the remaining 135 samples were divided into the calibration set (n = 90) and the prediction set (n = 45). Successive projections algorithm (SPA), genetic algorithm partial least squares (GAPLS) combined with SPA, weighted regression coefficient (Bw) and competitive adaptive reweighted sampling (CARS) were applied to select 14, 17, 24 and 25 effective wavelengths, respectively. Principal component analysis (PCA) and wavelet transform (WT) were applied to extract feature information with 20 and 58 features, respectively. PLS models were built based on the full spectra, the effective wavelengths and the features, respectively. All PLS models obtained good results. PLS models using full-spectra and features extracted by WT obtained the best results with correlation coefficient of calibration (r(c)) and correlation coefficient of prediction (r(p)) over 0.9. The overall results indicated that hyperspectral imaging combined with feature extraction methods could be used for detection of SSC in strawberry.

Visible/near-infrared spectroscopy was applied to determine the content of catalase (CAT) and peroxidase (POD) in barley leaves under the herbicide stress of propyl 4-(2-(4, 6-dimethoxypyrimidin-2-yloxy) benzylamino) benzoate (ZJ0273). The spectral data of the barley leaves in the range of 500-900 nm were preprocessed by moving average with 11 points. Seven outlier samples for CAT and 8 outlier samples for POD were detected and removed by Monte Carlo-partial least squares (MCPLS). PLS, least-squares support vector machine (LS-SVM) and extreme learning machine (ELM) models were built for both CAT and POD. ELM model obtained best results for CAT, with correlation coefficient of calibration (Rc) of 0.916 and correlation co-efficient of prediction (Rp) of 0.786. PLS model obtained best prediction results for POD, with Rc of 0.984 and Rp of 0.876. Successive projections algorithm (SPA) was applied to select 8 and 19 effective wavelengths for CAT and POD, respectively. PLS, LS-SVM and ELM models were built using the selected effective wavelengths of CAT and POD. ELM model performed best for CAT and POD prediction, with Rc of 0.928 and Rp of 0.790 for CAT and Rp of 0.965 and Rp of 0.941 for POD. The prediction results using the full spectral data and the effective wavelengths were quite close, and the prediction performance for POD was much better than the prediction performance for CAT, and the studies should be further explored to build more precise and more robust models for CAT and POD determination. The overall results indicated that it was feasible to use visible/near-infrared spectroscopy for CAT and POD content determination in barley leaves under the stress of ZJ0273.

The variety of Chinese cabbage seeds were recognized using hyperspectral imaging with 256 bands from 874 to 1,734 nm in the present paper. A total of 239 Chinese cabbage seed samples including 8 varieties were acquired by hyperspectral image system, 158 for calibration and the rest 81 for validation. A region of 15 pixel x 15 pixel was selected as region of interest (ROI) and the average spectral information of ROI was obtained as sample spectral information. Multiplicative scatter correction was selected as pretreatment method to reduce the noise of spectrum. The performance of four classification algorithms including Ada-boost algorithm, extreme learning machine (ELM), random forest (RF) and support vector machine (SVM) were examined in this study. In order to simplify the input variables, 10 effective wavelengths (EMS) including 1,002, 1,005, 1,015, 1,019, 1,022, 1,103, 1,106, 1,167, 1,237 and 1,409 nm were selected by analysis of variable load distribution in PLS model. The reflectance of effective wavelengths was taken as the input variables to build effective wavelengths based models. The results indicated that the classification accuracy of the four models based on full-spectral were over 90%, the optimal models were extreme learning machine and random forest, and the classification accuracy achieved 100%. The classification accuracy of effective wavelengths based models declined slightly but the input variables compressed greatly, the efficiency of data processing was improved, and the classification accuracy of EW-ELM model achieved 100%. ELM performed well both in full-spectral model and in effective wavelength based model in this study, it was proven to be a useful tool for spectral analysis. So rapid and nondestructive recognition of Chinese cabbage seeds by hyperspectral imaging combined with machine learning is feasible, and it provides a new method for on line batch variety recognition of Chinese cabbage seeds.

The feasibility of protein determination of shiitake mushroom (Lentinus edodes) using mid-infrared spectroscopy (MIR) was studied in the present paper. Wavenumbers 3 581-689 cm(-1) were used for quantitative analysis of protein content after removing of the part of obvious noises. Five points Savitzky-Golay smoothing was applied to pretreat the MIR spectra and partial least squares (PLS) model was built based on the pretreated spectra. The full spectra PLS model obtained poor performance with the ratio of prediction to deviation (RPD) of only 1.77. Successive projections algorithm (SPA) was applied to select 7 sensitive wavenumbers from the full spectra, and PLS model, multiple linear regression (MLR), back-propagation neural network (BPNN) and extreme learning machine (ELM) model were built using the selected sensitive wavenumbers. SPA-PLS model and SPA-MLR model obtained relatively worse results than SPA-BPNN model and SPA-ELM model. SPA-ELM obtained the best results with correlation coefficient of prediction (R(p)) of 0.899 5, root mean square error of prediction (RMSEP) of 1.431 3 and RPD of 2.18. The overall results indicated that MIR combined with chemometrics could be used for protein content determination of shiitake mushroom, and SPA could select sensitive wavenumbers to build more accurate models instead of the full spectra.

Near-infrared spectroscopy combined with chemometrics was used to investigate the feasibility of identifying different brands of soymilk powder and the counterfeit soymilk powder products. For this purpose, partial least squares-discriminant analysis (PLS-DA), linear discriminant analysis (LDA) and back-propagation neural network (BPNN) were employed as pattern recognition methods to class ify soymilk powder samples. The performances of different pretreatments of raw spectra were also compared by PLS-DA. PLS-DA models based on De-trending and multiplicative scatter correction (MSC)combined with De-trending(MSC+De-trending) spectra obtained best results with 100% prediction accuracy, respectively. Six and seven optimal wavenumbers selected by chi-loading weights of the best two PLS-DA models were used to build LDA and BPNN models. Results showed that BPNN performed best and correctly classified 100% of the soymilk powder samples for both the calibration and the prediction set. The overall results indicated that NIR spectroscopy could accurately identify branded and counterfeit soymilk powder products.

In the present study, hyperspectral imaging combined with chemometrics was successfully proposed to identify different varieties of black bean. The varieties of black bean were defined based on the three different colors of the bean core. The hy-perspectral images in the spectral range of 380-1,030 nm of black bean were acquired using the developed hyperspectral imaging system, and the reflectance spectra were extracted from the region of interest (ROD) in the images. The average spectrum of a ROI of the sample in the images was used to represent the spectrum of the sample and build classification models. In total, 180 spectra of 180 samples were extracted. The wavelengths from 440 to 943 nm were used for analysis after the removal of the spec- tral region with absolute noises, and 440-943 nm spectra were preprocessed by multiplicative scatter correction (MSC). Five classification methods, including partial least squares discriminant analysis (PLS-DA), soft independent modeling of class analogy (SIMCA), K-nearest neighbor algorithm (KNN), support vector machine (SVM) and extreme learning machine (ELM), were used to build discriminant models using the preprocessed full spectra, the feature information extracted by principal component analysis (PCA) and the feature information extracted by wavelet transform (WT) from the preprocessed spectra, respectively. Among all the classification models using the preprocessed full spectra, ELM models obtained the best performance; among all the classification models using the feature information extracted from the preprocessed spectra by PCA, ELM model also obtained the best classification accuracy; and among all the classification models using the feature information extracted from the preprocessed spectra by WT, ELM models obtained the best classification performance with 100% accuracy in both the calibration set and the prediction set. Among all classification models, WT-ELM model obtained the best classification accuracy. The overall results indicated that it was feasible to identify black bean varieties nondestructively by using hyperspectral imaging, and WT could effectively extract feature information from spectra and ELM algorithm was effective to build high performance classification models.

In the present study, Mid-infrared spectroscopy was used to identify the producing area of Letinus edodes, and relevance vector machine (RVM) was put forward to build classification models as a novel classification technique, and they obtained good performances. The head and the tail of the acquired mid-infrared spectra with the absolute noise were cut off, and the remaining spectra in the range of 3,581-689 cm(-1) (full spectra) of Letinus edodes were preprocessed by multiplicative scatter correction (MSC). Five classification techniques, including partial least Squares-discriminant analysis (PLS-DA), soft independent modeling of class analogy (SIMCA), K-nearest neighbor algorithm (KNN), support vector machine (SVM) and RVM, were applied to build classification models based on the preprocessed full spectra. All classification models obtained classification accuracy over 80%, KNN, SVM and RVM models based on full spectra obtained similar and good performances with classification accuracy over 90% in both the calibration set and the prediction set. The weighted regression coefficients (Bw) were used to select effective wave numbers of mid-infrared spectra and 6 effective wave numbers in total were selected on the basis of the weighted regression coefficients of PLS-DA model based on full spectra. PLS-DA, KNN, SVM and RVM models were built using these effective wave numbers. Compared with the classification models based on full spectra, PLS-DA models based on effective wave numbers obtained relatively worse results with classification accuracy less than 80%, and KNN, SVM and RVM obtained similar results in both calibration set and prediction set with classification accuracy over 90%. RVM performed well with classification rate over 90% based on full spectra and effective wave numbers. The overall results indicated that producing area of Letinus edodes could be identified by mid-infrared spectroscopy, while wave number selection and the RVM algorithm could be effectively used in mid-infrared spectroscopy analysis. In this study, mid-infrared spectroscopy was successfully applied to identify the producing area of Letinus edodes, which could provide a new concept for quality analysis of Letinus edodes and other agricultural products, and the application of mid-infrared spectroscopy had practical significance.

An identification method based on sparse representation (SR) combined with autoencoder network (AN) manifold learning was proposed for discriminating the varieties of transmission fluid by using near infrared (NIR) spectroscopy technology. NIR transmittance spectra from 600 to 1 800 nm were collected from 300 transmission fluid samples of five varieties (each variety consists of 60 samples). For each variety, 30 samples were randomly selected as training set (totally 150 samples), and the rest 30 ones as testing set (totally 150 samples). Autoencoder network manifold learning was applied to obtain the characteristic information in the 600-1800 nm spectra and the number of characteristics was reduced to 10. Principal component analysis (PCA) was applied to extract several relevant variables to represent the useful information of spectral variables. All of the training samples made up a data dictionary of the sparse representation (SR). Then the transmission fluid variety identification problem was reduced to the problem as how to represent the testing samples from the data dictionary (training samples data). The identification result thus could be achieved by solving the L-1 norm-based optimization problem. We compared the effectiveness of the proposed method with that of linear discriminant analysis (LDA), least squares support vector machine (LS-SVM) and sparse representation (SR) using the relevant variables selected by principal component analysis (PCA) and AN. Experimental results demonstrated that the overall identification accuracy of the proposed method for the five transmission fluid varieties was 97.33% by AN-SR, which was significantly higher than that of LDA or LS-SVM. Therefore, the proposed method can provide a new effective method for identification of transmission fluid variety.

The potential of hyperspectral imaging technology was evaluated for discriminating three types of waxed apples. Three types of apples smeared with fruit wax, with industrial wax, and not waxed respectively were imaged by a hyperspectral imaging system with a spectral range of 308-1 024 nm. ENVI software processing platform was used for extracting hyperspectral image object of diffuse reflection spectral response characteristics. Eighty four of 126 apple samples were selected randomly as calibration set and the rest were prediction set. After different preprocess, the related mathematical models were established by using the partial least squares (PLS), the least squares support vector machine (LS-SVM) and BP neural network methods and so on. The results showed that the model of MSC-SPA-LSSVM was the best to discriminate three kinds of waxed apples with 100%, 100% and 92.86% correct prediction respectively.

Early diagnosis of gray mold on tomato stalks based on hyperspectral data was studied in the present paper. A total of 112 samples' hyperspectral data were collected by hyperspectral imaging system. The study spectral region was from 400 to 1,030 nm. Combined with image processing and chemometric methods, the tomato stalk gray mold diagnosis models were built. Seven effective wavelengths were selected by analysis of variable load distribution in PLS model. The experimental results showed that the excellent results were achieved by EW-LS-SVM model with standard normal variate (SNV) spectral and multiplicative scatter correction (MSC) spectral, and the accuracy of diagnosing gray mold on tomato stalks was satisfied and better than PLS model with whole band. Hence, it is feasible to early diagnose gray mold on tomato stalks using hyperspectral imaging technology, which provides a new early diagnosis and warning method for tomato disease.

Near infrared (NIR) spectroscopy was applied for the fast and nondestructive determination of malondialdehyde (MDA) content in oilseed rape leaves. A total of 90 leaf samples were collected, the calibration set was composed of 60 samples, and the prediction set was composed of 30 samples. Different preprocessing methods were used before the calibration stage, including smoothing, standard normal variate, first and second derivative, and detrending. Then partial least squares (PLS) models were developed for the prediction of MDA content in oilseed rape leaves. The latent variables selected by PLS and effective wavelengths selected by successive projections algorithm (SPA) were used as the inputs of least square-support vector machine (LS-SVM) to develop LV-LS-SVM and SPA-LS-SVM models. The correlation coefficients (r) and root mean square error of prediction (RMSEP) were used as the model evaluation indices. Excellent results were achieved by LV-LS-SVM model, and the prediction results by LS-SVM model using detrending spectra were r = 0.999 9 and RMSEP = 0.530 2, and those by LS-SVM model using 2-Der spectra were r = 0.999 9 and RMSEP = 0.395 7. The results showed that NIR spectroscopy could be used for determination of MDA content in oilseed rape leaves, and an excellent prediction precision was achieved. This study supplied a new approach to the dynamic and continuous field monitoring of growing status of oilseed rape.