**3.4. Viral metagenomics**

NGS has revolutionized metagenomics in a major way by ensuring high data throughput and by removing the hassles of cultivation/isolation by providing cost-effective options. Metage‐ nomics involves sequencing of samples from diverse environments spanning across the biosphere [64]. The initial attempts at characterizing the viral metagenomes were more of an enumeration nature [65] and provided a glimpse of the enormous diversity underlying the previously unculturable communities. NGS has paved way for extensive characterization of the functional role of virome in hosts harbouring them [66, 67]. Analysis of metagenomics data is challenging as it includes simultaneous assembly of multiple genomes/transcriptomes and the complex interplay between them. Two major methods based on 'sequence-similarity' and 'sequence composition' are usually used for categorization of samples in metagenomics. It has been observed that the alignment-free 'sequence composition'-based methods provide better means of classifying viral samples as 'sequence similarity'-based methods could only classify up to 30% of the reads [68].

In a major study involving analysis of dsDNA viruses from 43 ocean samples obtained from across the globe revealed several intriguing observations [69]. Genes shared across different samples were used as 'core genes' for comparison. 'Niche-differentiation' of different viral populations based on the layer of the ocean they occupy was observed. As viruses rely on the host machinery to replicate, a direct relationship was observed between the community structures of both viruses and hosts. Environmental factors like salinity also influenced the viral persistence and hence their diversity. Technological advances in viral metagenomics would help to unravel the underlying rules of viral evolution and ecology, the so-called 'Genomic rulebook of viruses' [70].
