Host Taxonomy Discrepancy: GTDB vs NCBI in Phanta Database

[bablukr056]

Dear Author

I have been using the UHGV pipeline for host predictions, and my host taxonomy is based on GTDB (https://github.com/snayfach/UHGV: "[GTDB r207 and GTDB-tk were used to assign taxonomy to all prokaryotic genomes") . However, I noticed that the taxonomy in the Phanta database
taxonomy for Prokaryotic Portion (HumGut + UHGV "MQ+") is based on NCBI taxonomy.

Request:
Could you please guide me on how to align or convert the host taxonomy from the Phanta database (NCBI) with GTDB? Alternatively, is there a way to adjust my workflow or the database to ensure consistency in taxonomy when performing host predictions?

Thank you

[yipinto]

Hi Bablu,
Humgut has both gtdb and ncbi taxonomy see here, so I think you should be easily convert from NCBI to GTDB for the prokaryotic portion.
However, maybe the easiest solution for you will be to use a Phanta database that include GTDB taxonomy?
For example UHGGv2 + UHGV "MQ+"
See all databases here
Yishay

[bablukr056]

Hi Yishay,

Thank you for your valuable suggestion and reply.

Could you share any methods, tools, or scripts for converting from NCBI to GTDB? I’d appreciate any insights you might have.

For my study, I am specifically using HumGut, and I plan to incorporate the UHGGV2 database at a later stage.

Additionally, I’d like to suggest the possibility of integrating HumGut + UHGV "MQ+" with the GTDB prokaryotic section. This could be beneficial not only for me but also for other users.

Looking forward to your thoughts.

Best,
Bablu

[yipinto]

Thanks for your suggestion.
See the HumGut.tsv, it includes both gtdb taxonomy and ncbi taxonomy so you can convert.