JINDEX: JSON and index search system for plant germplasm database

Abstract

To facilitate the development of new varieties and conserve plant germplasm data for future needs, Thailand organizations have developed plant germplasm search systems that provide access and exchange plant genetic resource data like the international organizations. However, the development of plant germplasm search systems using traditional data warehousing creates a limitation in terms of structural flexibility and scalability, as well as search performance. Accordingly, this paper proposes a JSON and index search system for the plant germplasm database named JINDEX. JINDEX is a new approach for implementing a galaxy schema in NoSQL based a hybrid key-value/document data model. Plant germplasm data is stored in JSON files as a document data model, providing a flexible structure of genetic resources for various plants. The plant germplasm data and index files are organized as an on-disk tree structure, which can be accessed regarding a key-value data model to retrieve plant germplasm data efficiently. Experimental results showed that the JINDEX-based plant germplasm search system outperforms the previous plant germplasm search system implemented in a relational database in terms of query response time. Additionally, this JINDEX-based plant germplasm search system has been deployed since 2020, demonstrating that it can work practically.