Day 3: Classification and Taxonomy Practical

Phage classification and taxonomy

1️⃣ Finding the closest relatives for your phage

Using NCBI Virus sequence searches

When you only have a few genomes, and you are most comfortable with web-based searches, we can recommend NCBI Virus.

You can search for relatives with the Search by sequence function using the whole genome (nucleotide search) or a single protein. The underlying algorithm is the same BLAST you have probably used before.

💡 Try searching with the genome sequence first. If this yields results with both high % coverage and % identity values, then you have close relatives. If not, you can use searches with proteins to find more distant relatives.

My search using the restarted T7 genome yields 499 results. The first thing to do is to sort them according to % coverage (click on the column header) so that we can see the database sequences that share the highest fraction of their genome with our search query. Now we can see that the top "hits" are all isolates belonging to the same species as T7. From this list you can select the closest relatives for further inspection or download. We suggest not to download more than 20 sequences for the next step.

Tips

• Use the Refine Search button to narrow down the number of search results shown
• Limit to Sequence type > RefSeq if you only want one representative of each species
• Use the select column to limit or expand the information shown

Further investigations within NCBI Virus

Once you have selected a subset of sequences, you can use the Align and Build phylogenetic tree buttons to further investigate the search results. However, this will not include your own phage sequence.

⚠️ If you select too many genomes, these features will not work.

Using NCBI Virus taxonomy information

If you have done the investigation of your genome in the previous step, you may have already found information on closest relatives from the prokka information. You can also use the NCBI virus interface to search for a specific taxon.

For example, the search for Tequatrovirus gives 1497 nucleotide sequences, of which 251 are flagged as complete and 83 RefSeq records.

Using VipTree to find closest relatives

If your BLAST searches yield little results, you can alternatively submit your genome to the VipTree server which will compare the translated genome with the reference genomes in its database and add it to the tree.

💡 VipTree has a lot of features that you can explore on your own time.

2️⃣ Assessment of the intergenomic distances