Schematic diagram of the bioinformatic analysis workflow. (1) Viral RdRp protein sequences were identified using the HMM search-and-update pipeline, which assisted the identification of highly divergent RdRp amino acid sequences. The dashed arrows represent the two iterations of this process. Subsequently, true-positive RdRps were obtained using PalmScan and WebLogo. (2) Taxonomic assignment of RdRps based on the ICTV classification. (3) RdRp nucleotide sequences were clustered using 95% ANI and 85% AF to obtain the vOTUs. (4) AAI distribution histograms were calculated using EzAAI. (5) Phylogenetic analysis was performed for those vOTUs that contained two virus species.