Identification and characterization of tissue- and stress-specific circular RNAs (circRNAs) of tea to generate the largest tea circRNAs data repository

Enormous RNA-seq data of tea (Camellia sinensis) is available at NCBI, yet only 3,174 circular RNAs (circRNAs) have been reported to date. This study establishes a foundation for using public datasets to generate circRNA reservoirs. After standardizing the pipeline using in-house RNA-seq data from Indian tea, 178 RNA-seq datasets of tea were downloaded from NCBI and used for the generation of the largest tissue and stress-specific tea circRNAs repository (TCDB; http://indianteagenome.in:8080/tcdb/ ), till date. A total of 3052 and 59,575 full-length circRNA isoforms were identified from Indian tea and downloaded public datasets, respectively. Notably, only 10% of these shared homologies with known plant circRNAs, indicating that 90% are novel. Since only 9.77% circRNAs were shared between 25,651 and 39,750 circRNAs from stress- and tissue-related RNA-seq data, respectively, mostly are either tissue- or stress-specific. Annotation of parental genes of differentially expressed circRNAs (DECs) highlighted their roles in important pathways, including phenylpropanoid biosynthesis, aminoacyl-tRNA biosynthesis, and regulation of different metabolisms. Although circRNA-miRNA interaction analysis favors the role of circRNAs as miRNA sponges, with 17% of circRNAs having no miRNA interactions, there could be other possible functions of circRNAs. While circRNA data generated at this large scale could serve as foundation for in-depth studies on tea circRNAs, the strategy of using available public RNA-seq data to detect circRNAs and conduct further downstream analysis will aid in accelerating similar studies in other species.