Yield prediction is the primary goal of genomic selection (GS)-assisted crop breeding. Because yield is a complex quantitative trait, making predictions from genotypic data is challenging. Transfer learning can produce an effective model for a target task by leveraging knowledge from a different, but related, source domain and is considered a great potential method for improving yield prediction by integrating multi-trait data. However, it has not previously been applied to genotype-to-phenotype prediction owing to the lack of an efficient implementation framework. We therefore developed TrG2P, a transfer-learning-based framework. TrG2P first employs convolutional neural networks (CNN) to train models using non-yield-trait phenotypic and genotypic data, thus obtaining pre-trained models. Subsequently, the convolutional layer parameters from these pre-trained models are transferred to the yield prediction task, and the fully connected layers are retrained, thus obtaining fine-tuned models. Finally, the convolutional layer and the first fully connected layer of the fine-tuned models are fused, and the last fully connected layer is trained to enhance prediction performance. We applied TrG2P to five sets of genotypic and phenotypic data from maize (Zea mays), rice (Oryza sativa), and wheat (Triticum aestivum) and compared its model precision to that of seven other popular GS tools: ridge regression best linear unbiased prediction (rrBLUP), random forest, support vector regression, light gradient boosting machine (LightGBM), CNN, DeepGS, and deep neural network for genomic prediction (DNNGP). TrG2P improved the accuracy of yield prediction by 39.9%, 6.8%, and 1.8% in rice, maize, and wheat, respectively, compared with predictions generated by the best-performing comparison model. Our work therefore demonstrates that transfer learning is an effective strategy for improving yield prediction by integrating information from non-yield-trait data. We attribute its enhanced prediction accuracy to the valuable information available from traits associated with yield and to training dataset augmentation. The Python implementation of TrG2P is available at https://github.com/lijinlong1991/TrG2P. The web-based tool is available at http://trg2p.ebreed.cn:81.

How to evaluate the accuracy of quantitative trait prediction is crucial to choose the best model among several possible choices in plant breeding. Pearson’s correlation coefficient (PCC), serving as a metric for quantifying the strength of the linear association between two variables, is widely used to evaluate the accuracy of the quantitative trait prediction models, and generally performs well in most circumstances. However, PCC may not always offer a comprehensive view of predictive accuracy, especially in cases involving nonlinear relationships or complex dependencies in machine learning-based methods. It has been found that many papers on quantitative trait prediction solely use PCC as a single metric to evaluate the accuracy of their models, which is insufficient and limited from a formal perspective. This study addresses this crucial issue by presenting a typical example and conducting a comparative analysis of PCC and nine other evaluation metrics using four traditional methods and four machine learning-based methods, thereby contributing to the improvement of practical applicability and reliability of plant quantitative trait prediction models. It is recommended to employ PCC in conjunction with other evaluation metrics in a targeted manner based on specific application scenarios to reduce the likelihood of drawing misleading conclusions.

Computer vision and image processing technology have been rapidly developed and widely applied in many fields. There are many potential applications in modern agriculture. In this paper, a novel vegetable disease and insect pest recognition method is proposed based on the current computer vision and image processing methods. To investigate the vegetable disease and insect pest state, it is convenient to use images captured using smart phones for judgment. To implement this application, the disease area and the insect number on the leaves should be detected and figured out. So a new extraction and classification algorithm is firstly introduced to recognize leaves from images. Then a region-labeling algorithm is applied to calculate the insect number and disease areas in the segmented images. To deal with the areas of adhesion, a mathematical morphology algorithm is used for separating the objects. The proposed method is implemented on mobile smart devices and tested with field experiments. The experimental results show that the proposed method has good recognition performance with high efficiency. © 2013 Copyright TSI® Press.

Farm Management Information Systems (FMISs) are being expanded to improve operation efficiency, reduce inputs, and ensure compliance with standards and regulations. However, this goal is difficult to attain in the vegetable sector, where data acquisition is time-consuming and data at different stages is fragmented by the potential diversity of crops and multiple batches cultivated at any given farm. This applies, in particular, to farms in China, which have small areas and low degrees of mechanization. This study presents an integrated approach to track and trace production efficiently through our Digital Farm Management System (DFMS), which adopts the cloud framework and utilizes Quick Response (QR) codes and Radio Frequency Identification (RFID) technology. Specifically, a data acquisition system is proposed that runs on a smartphone for the efficient gathering of planting information in the field. Moreover, DFMS generates statistics and analyses of planting areas, costs, and yields. DFMS meets the FMIS requirements and provides the accurate tracking and tracing of the production for each batch in an efficient manner. The system has been applied in a large-scale vegetable production enterprise, consisting of 12 farms distributed throughout China. This application shows that DFMS is a highly efficient solution for precise vegetable farm management.

GAHP is a freely available software package for genetic analysis with bi-parental immortalized heterozygous and pure-line populations. The package is project-based and integrated with multiple functions. All operations and running results are properly saved in a project, which can be recovered when the project is re-open by the package. Four functionalities have been implemented in the current version of GAHP, i.e., 1) MHP: visualization of genetic linkage maps; 2) VHP: analysis of variance (ANOVA) and estimation of heritability on phenotypic data; 3) QHP: quantitative trait locus (QTL) mapping on both genotypic and phenotypic data; 4) SHP: simulation of bi-parental immortalized heterozygous and pure-line populations, and power analysis of QTL mapping. VHP and QHP can be conducted in individual populations, as well as in multiple populations by the combined analysis. Input files are arranged either in the plain text format with an extension name same as the functionality or in the MS Excel formats. Output files have the same prefix name as the input file, but with different extensions to indicate their contents. Three characters before the extension names stand for the types of populations used in analysis. In the interface of the software package, input files are grouped by functionality, and output files are grouped by individual or combined mapping populations. In addition to the text-format outputs, the constructed linkage map can be visualized per chromosome or for a number of selected chromosomes; line plots and bi-plots can be drawn from QTL mapping results and phenotypic data. Functionalities and analysis methods available in GAHP help the investigation of genetic architectures of complex traits and the mechanism of heterosis in plants.

Combining ability analysis can be used to preliminarily identify the advantages and disadvantages of combinations and parents in earlier generations, enabling breeders to reduce the range of material, save breeding time, and improve breeding efficiency. An approach for combining ability analysis through the hybrid crop breeding information management system is presented. The general combining ability prediction effect of parents and the specific combining ability prediction effect of combinations are calculated to analyze hybrid combinations using the hybrid crop breeding information management system. The results provide the basis for parent selection and combination selection. The plant breeding trial management function of the system can provide convenient diallel crossing trial design, field planting plan, and combining ability analysis. In the system, the genealogy of breeding materials is traced with the combining ability test crosses. The selection of high-generation breeding materials can be performed in accordance with the combining ability test results of early generation materials. The system has been successfully applied to a large Chinese seed company. The combining ability test function automates data analysis and eliminates days in the decision-making process. The efficiency of the combining ability test analysis and test report generation has improved to more than double by using the system.