A natural ANI gap that can define intra-species units of bacteriophages and other viruses

Despite the importance of intra-species variants of viruses for causing disease and/or disrupting ecosystem functioning, there is no universally applicable standard to define these. A 95% whole-genome average nucleotide identity (ANI) gap is commonly used to define species, especially for bacteriophages, but whether a similar gap exists within species that can be used to define intra-species units has not been evaluated yet. Whole-genome comparisons among members of 1,016 bacteriophage species revealed a region of low frequency of pairs around 99.2-99.8% ANI, showing 3-fold or fewer pairs than expected for an even or normal distribution. This second gap is prevalent in viruses infecting various cultured or uncultured hosts, and from a variety of environments, although a few exceptions to this pattern were also observed ([~]3.7% of the total species evaluated) and are likely attributed to cultivation biases. Similar results were observed for a limited set of eukaryotic viruses that are adequately sampled including SARS-CoV-2, whose ANI-based clusters matched well the WHO-defined Variants of Concern, indicating that they represent functionally and/or ecologically distinct units. The existence of sequence-discrete units appears to be predominantly driven by (high) ecological cohesiveness coupled to either recombination frequency for bacteriophages or selection and clonal evolution for other viruses such as SARS-CoV-2. These results indicate that fundamentally different underlying mechanisms could lead to similar diversity patterns. Based on these results, we propose the 99.5% ANI as a practical, standardized, and data-supported threshold for defining viral intra-species units of bacteriophages, for which we propose the term genomovars. ImportanceViral species are composed of an ensemble of intra-species variants whose dynamic may have major implications for human and animal health and/or ecosystem functioning. However, the lack of universally-accepted standards to define these intra-species variants has led researchers to use different approaches for this task, creating inconsistent intra-species units across different viral families and confusion in communication. By comparing hundreds of viral bacteriophage genomes, we show that there is a nearly universal natural gap in whole-genome average nucleotide identities (ANI) among genomes at around 99.5%, which can be used to define intra-species units. Therefore, these results advance the molecular toolbox for tracking viral intra-species units and should facilitate future epidemiological and environmental studies.