Deepika Sharma 1*, Aishwarya Nayar 2 and Ashutosh Sharma 3
1Amity Institute of Organic Agriculture, Amity University, Noida 201303, India
2Department of Plant Pathology, Dr. Yashwant Singh Parmar University of Horticulture and Forestry,
Nauni, Solan 173230, India
3Faculty of Agricultural Sciences, DAV University, Jalandhar 144012, India
*(e-mail: dpkasharma44@gmail.com)
(Received: 24 February 2025; Accepted: 6 June 2025)
ABSTRACT
Plant viruses and viroids cause extensive losses with reduction in crop productivity worldwide. The emergence of high-throughput sequencing technologies, commonly referred to as ‘Next-generation sequencing’ together with the metagenomics approach has led to a rapid increase in our understanding of plant viral communities. The utilization of high throughput NGS technologies has proven to be effective in the detection of previously unidentified disease-associated with new pathogens including viruses. Virome analysis using high-throughput sequencing technologies leads to the exploration of different viruses. These technologies, in combination with automation, artificial intelligence can allow for the efficient utilization of plant disease clinics in virome diagnostics. High-throughput sequencing methods have advantages of identification and genomic characterization of viruses and are important for diversity studies of plant viromes. Plant virome studies have the capability to carry out the detection of unknown viruses in mixed infection to reveal the presence of novel viruses. Further, the new machine learning/deep learning tools have enabled the detection of new viral sequences in already available host nucleotide sequences, enabling us to identify lysogenic viruses. In the era of metagenomics, plant-specific virome studies will help in checking the potential epiphytotic soon. Therefore, the present review highlights the successful utilization of high-throughput sequencing technologies in characterizing plant virome.
Key words : metagenomics, next generation sequencing, viral communities, deep learning