Blast is not returning taxonomy results: Fetch Taxonomic Ranks

usegalaxy.eu support
, , ,
1

Dear Galaxy,

I have been trying to run blast+ blast n with taxonomy results for scientific and comon name against the NCBI NT (2025), but it returns N/A on those fields. Tested a re-run on files from last week where it worked fine, and the same thing occurs. Bellow is an example, the exact same job ran last week (first picture) and run today (second picture). Could you help with this issue?

image
image423Ɨ705 30.5 KB

image
image292Ɨ674 15.3 KB

Kind regards

2

Welcome @SAlves

Interesting!

Would you be able to confirm a few items so I can look into this closer and get it reported?

  1. Server: this was at UseGalaxy.eu correct? The server had some tools updated recently, so this will be part of the review.
  2. Tool version: BLASTN+ but the wrappers have a versioning. Find this is on the rerun forms at the top or review on the job details page (i-icon). My guess is that these data were run using different versions but letā€™s confirm so I can reproduce exactly.
  3. Database index version: NCBI NT (2025) for both it seems but please double check.

Please capture these details and let us know. If you want to post back the job details views in a screenshot that would be Ok, or you can share the history with the runs, but I think just the core details above will be enough for this.

What to do: if you want to reproduce results, try using the tool version that has the taxonomy attached. See ā†’ FAQ: Changing the tool version

Thank you for reporting the issue and Iā€™ll watch for your reply! :slight_smile:

Shovill producing blank contig graph file: Use Galaxy Version 1.1.0+galaxy2
3

Hello @jennaj,

Thank you for your reply.

I confirm it is the EU server. The Database index version I need to correct myself for both runs is NCBI NT (2024).

image

The tool ID version from today and last week seem to be the same but I noticed it returns an error that didnā€™t last week, see the pictures below.

Today:

image

And from last week:

image

I noticed this because I had some new files I wanted to run with a saved workflow, and the following steps require the taxonomy information for some data ā€œcleanupā€. Thankfully I had this data still saved for the test :slightly_smiling_face: .

Please let me know if you require more information.

4

Hi @SAlves

Thanks for sharing the extra details! Very helpful.

I can replicate your result. The issue is with a different tool that is commonly used after mapping to pull in taxonomy information. Your workflow likely includes it but please check to make sure we are capturing all of the related functions that may be impacted by the missing index.

Thanks for reporting the issue! This is important to correct so letā€™s leave this topic open until resolved. :hammer_and_wrench:

5

The error: ā€œThe -taxids command line option requires additional data filesā€ likely indicates that these 3 files are missing in $BLASTDB:
18M taxdb.bti
164M taxdb.btd
83M taxonomy4blast.sqlite3
Also ā€“ ā€˜ntā€™ dated 2024 is both old and includes an excessive #sequences. NCBI now recommends using ā€˜core_ntā€™ blastdb.
The Galaxy admin can download the 3 taxonomy files :
cd $BLASTDB; update_blastdb.pl --source gcp taxdb --decompress

1 Like
6

The other tool just filters results based on the taxonomy results from blast, and since blast is not returning anything it canā€™t filter the table.

7

Hi @Eugene_Spier

Yes, there are two things that could be changed (in addition to the original tool correction). Letā€™s summarize:

Repair the taxonomy fetching tool

I just check and it is not resolved yet. Iā€™ll follow up more about this. Comments will post on the ticket for intermediate updates.

Blast indexes and form options ā€“ whatā€™s supported.

The extended taxonomy options have never been supported (intentionally) so far at any of the public Galaxy servers, and the message you noticed in the logs is expected if this is attempted. There were two technical reasons but Iā€™m not certain if both are still relevant or what else could be done to make it happen. Iā€™ll also follow up about this to get some clarity.

Update Blast indexes ā€“ agree!

I added in the other database to an existing ticket. This is enough to prompt the whole package to be updated. Iā€™ve contributed to these updates in the past, and maybe can help again this time if someone else canā€™t pick it up. More feedback will be added to the ticket, plus Iā€™ll be updating back at this topic as each of these items is reviewed.

Then for your comment @SAlves

The taxonomy fetching tool works by using the subject GI identifiers for a look up. The target information is always included in outputs (unless specifically omitted?), and when the target is a database from NCBI, all of the artifacts will have a GI number available.

While the GI is not included in the default 25 tables, this can be toggle into an output in the extended tabular options as sgi = Subject GI. Maybe this helps you? I have an example in my testing history linked to the Github ticket for the taxonomy tool correction above. Please ask if you need to do this and are not sure how ā€“ sharing a screenshot of your current options would be a good place to start!

More soon! :slight_smile:

8

Just the GI is not enough for what I am working on, but grouping them by scientific name and common name worked 3 weeks ago, so Iā€™m guessing as soon as the indexing thing is corrected I can rerun the workflow and will work as expected. Hope it gets fixed soon.

Thank you for your help :smiling_face:.

A terƧa, 24/02/2026, 17:49, Jennifer Hillman-Jackson <notifications@galaxy.discoursemail.com> escreveu:

1 Like
9

Yes, the GI is just the relational key connection and is probably what is included in your workflow already (if it was run at one of the public servers). Iā€™m following up to see what can be done for the taxonomy tool correction (root issue to restore prior functionality).