AAI for viruses

[SergeyBaikal]

Dear colleagues, have you tried optimizing the parameters of the program for calculating AAI for viral genomes? To what extent is this possible?

[zhangwenda0518]

I have the same question, is ezaai suitable for virus data?

[endixk]

I don't expect any specific calibrations would be needed, given you have a decent genome data to compare against.

There are some practical use cases in viral community (listed below), seemingly without particular taxonomic restrictions.

• Riboviria RdRp study: https://academic.oup.com/ve/article/10/1/veae059/7720591
• Nucleocytoviricota genomes: https://link.springer.com/article/10.1007/s10811-025-03518-9
• Bacteriophage genomes: https://www.nature.com/articles/s41467-023-44370-0

But the results should be handled with care, especially if the clade is highly divergent, beyond the twilight zone of amino acid alignments. CDS extraction from nucleotide data would also not be very feasible since we employ Prodigal, which is optimized for prokaryotic sequences.