Convolutional neural network (CNN) can be used to quickly identify crop seed varieties. 1200 seeds of ten soybean varieties were selected, hyperspectral images of both the front and the back of the seeds were collected, and the reflectance of soybean was derived from the hyperspectral images. A total of 9600 images were obtained after data augmentation, and the images were divided into a training set, validation set, and test set with a 3:1:1 ratio. Pretrained models (AlexNet, ResNet18, Xception, InceptionV3, DenseNet201, and NASNetLarge) after fine-tuning were used for transfer training. The optimal CNN model for soybean seed variety identification was selected. Furthermore, the traditional machine learning models for soybean seed variety identification were established by using reflectance as input. The results show that the six models all achieved 91% accuracy in the validation set and achieved accuracy values of 90.6%, 94.5%, 95.4%, 95.6%, 96.8%, and 97.2%, respectively, in the test set. This method is better than the identification of soybean seed varieties based on hyperspectral reflectance. The experimental results support a novel method for identifying soybean seeds rapidly and accurately, and this method also provides a good reference for the identification of other crop seeds.

Hyperspectral imaging is a nondestructive testing technology that integrates spectroscopy and iconology technologies, which enables us to quickly obtain both internal and external information of objects and identify crop seed varieties. First, the hyperspectral images of ten soybean seed varieties were collected and the reflectance was obtained. Savitzky-Golay smoothing (SG), first derivative (FD), standard normal variate (SNV), fast Fourier transform (FFT), Hilbert transform (HT), and multiplicative scatter correction (MSC) spectral reflectance pretreatment methods were used. Then, the feature wavelengths and feature information of the pretreated spectral reflectance data were extracted using competitive adaptive reweighted sampling (CARS), the successive projections algorithm (SPA), and principal component analysis (PCA). Finally, 5 classifiers, Bayes, support vector machine (SVM), k-nearest neighbor (KNN), ensemble learning (EL), and artificial neural network (ANN), were used to identify seed varieties. The results showed that MSC-CARS-EL had the highest accuracy among the 90 combinations, with training set, test set, and 5-fold cross-validation accuracies of 100%, 100%, and 99.8%, respectively. Moreover, the contribution of spectral pretreatment to discrimination accuracy was higher than those of feature extraction and classifier selection. Pretreatment methods determined the range of the identification accuracy, feature-selective methods and classifiers only changed within this range. The experimental results provide a good reference for the identification of other crop seed varieties.

Low-phosphorus (low-P) stress has a significant limiting effect on crop yield and quality. Although the molecular mechanisms of the transcriptional level responsible for the low-P stress response have been studied in detail, the underlying epigenetic mechanisms in gene regulation remain largely unknown. In this study, we evaluated the changes in DNA methylation, gene expression and small interfering RNAs (siRNAs) abundance genome-wide in response to low-P stress in two representative soybean genotypes with different P-efficiencies. The DNA methylation levels were slightly higher under low-P stress in both genotypes. Integrative methylation and transcription analysis suggested a complex regulatory relationship between DNA methylation and gene expression that may be associated with the type, region, and extent of methylation. Association analysis of low-P-induced differential methylation and gene expression showed that transcriptional alterations of a small part of genes were associated with methylation changes. Dynamic methylation alterations in transposable element (TE) regions in the CHH methylation context correspond with changes in the amount of siRNA under low-P conditions, indicating an important role of siRNAs in modulating TE activity by guiding CHH methylation in TE regions. Together, these results could help to elucidate the epigenetic regulation mechanisms governing the responses of plants to abiotic stresses.

Phytocyanins (PCs), plant-specific blue copper proteins, are crucial for various biological processes during plant development. However, a comprehensive characterization of the soybean PC gene family (GmPC) is lacking. In this study, we performed genome-wide screening of soybean PC genes, and 90 PC genes were identified in the soybean genome. Further analysis revealed that the GmPC family was categorized into four subfamilies (stellacyanins, GmSCs; uclacyanins, GmUCs; plantacyanins, GmPLCs; and early nodulin-like proteins, GmENODLs). In-depth analysis revealed that each specific GmPC subfamily exhibited similar characteristics, with segmental duplications playing a major role in expanding the members of GmPC. Additionally, synteny and evolutionary constraint analyses suggested that GmPCs have undergone strong selective pressure for purification during the evolution of soybeans. The promoter cis-regulatory elements analysis of GmPCs suggested that GmPCs might play a crucial role in various stress responses. The expression patterns of GmPCs exhibited tissue-specific variations. Moreover, 23 of the GmPCs may be involved in soybean’s response to salt stress. In all, our study presents a systematic overview of GmPC, which not only provides a valuable foundation for further functional investigations of GmPCs, but also offers new insights into the mechanism of soybean salt tolerance.