Research On Big Data System For Crop Germplasm Resources

With the rapid development of biotechnology and multi-source data acquisition methods,a large number of structured,semi-structured and unstructured data have been generated in the research field of crop germplasm resources,and crop germplasm resources have entered the era of big data.Big data technology has absolute advantages in massive heterogeneous data processing.Through highly automated analysis of data,inductive reasoning can be made,and potential patterns and rules can be mined out to provide more valuable information for the society.Although some scholars in China have gradually realized the importance of building big data of crop germplasm resources in recent years,the related concepts of big data of crop germplasm resources need to be defined and clarified,and the big data system of crop germplasm resources and its contents need to be established and implemented.Therefore,based on the research status of big data of crop germplasm resources,this paper made some theoretical exploration and practical innovation on the construction of big data system of crop germplasm resources,aiming at providing theoretical basis and application reference for further promoting the construction of big data of crop germplasm resources.The research work of this paper mainly includes the following aspects:(1)The overall framework of big data system of crop germplasm resources was constructed.The basis of system construction based on big data and crop germplasm resources was put forward,eight system construction principles such as applicability,systematicness and compatibility were formulated,and the four-dimensional element structure of big data of crop germplasm resources with "foundation-goal-process-guarantee" as the main route was put forward,including operational requirements of crop germplasm resources,big data technology,data processing and standards,etc.The boundary and content of system construction were determined,and the overall framework of big data system of crop germplasm resources was designed.(2)The supporting system is designed in detail.Based on the overall framework of big data system of crop germplasm resources,the standard specification system,data system,management system and security system were designed in detail according to the design idea of first overall then local and progressive step by step,which enriched the content of the system.The framework of six types of standards and specifications for big data of crop germplasm resources was established,and the data classification,coding specifications,exchange and sharing of the data system were established.A multi-dimensional management system based on management team building,data management and system operation and maintenance was designed,and a security system was established to ensure the safe and reliable operation of the big data system.(3)Design the technology/tool subsystem.Based on the processing elements of big data of crop germplasm resources,the technical system was designed.Aiming at the data acquisition technology of crop germplasm resources,the technical scheme of secondary data acquisition was put forward,the pretreatment technology was defined,and the hybrid data storage technology scheme based on cloud technology was put forward.The analysis technical framework and visualization technical system suitable for crop germplasm resources data display were designed,and the platform/tool framework was designed in combination with the technical system.Taking the number of ears per unit area of wheat as an example,an empirical study was carried out,which improved the collection method of ear images,marked about 10,000 samples for training,and optimized the parameters of deep learning target detection algorithm.The results showed that the F1 evaluation index of the model was greater than 0.91,which was close to the practical effect,and verified the feasibility of technical system design.(4)The application subsystem is designed and the knowledge map is constructed.The construction principle of application subsystem was put forward,and the application subsystem was divided into three parts: basic work,basic research and applied basic research according to the business requirements of crop germplasm resources research and user groups,and the docking scheme between application system and technical system was designed.Taking the construction of knowledge map of crop germplasm resources as an example,the model layer of knowledge map centered on crop germplasm resources was designed,and the data layer of knowledge map was constructed with Protégé software.The data expansion method based on knowledge map was proposed,and the basic sociological data related to crop germplasm resources was obtained.The visualization interface of knowledge map based on graph structure was constructed,which verified the rationality of application system design.To sum up,this paper puts forward the basis,principles and elements of big data system construction of crop germplasm resources,designs the overall framework of big data system,and elaborates and analyzes the two core subsystems-technology/tool subsystem and application subsystem in the framework from three levels of theory,technology and practice,and clarifies the main body,application technology and development direction of big data construction of crop germplasm resources,which has important theoretical and application value for future big data construction of crop germplasm resources.