Kraken2 Database Addition GTDB v220

usegalaxy.eu support
1

Is there any possibility of adding the GTDB v220 database to Kraken2??? Also the EupathDB 46 would be helpful.

2

Hi @Jon_Colman

Both of these indexes are available at UseGalaxy.org for now. There are some technical reason why these are not at the EU server yet. When possible, all the UseGalaxy servers will host them.

GTDB

org-kraken2-gtdb
org-kraken2-gtdb1290Ă—482 35.7 KB

EuPathDB

org-kraken2-eupathdb
org-kraken2-eupathdb1204Ă—438 32.2 KB

I believe you know how to do this already but I’ll share for anyone else reading this later! Entire histories or just individual files can be moved between the servers by URL.

Hope this helps! :slight_smile:

3

Yes, I’ve been transferring over to run them on the main site. Of note, on usegalaxy.eu I would normally only use Kraken2 on the smaller databases, as the Core-nt would take forever. Seems to be working well now.

1 Like
4

Thanks for letting is know. And agree, core-nt jobs will take some time. The reference data is very large, then filtering the results takes time. I’m really glad we can host these! They would be unlikely to work on a laptop, or even most desktops. :slight_smile:

5

It was worse than that on the .eu, even files not very large using the core-nt database, I would start it and it would start running, even after a whole day it wouldn’t complete. Now it runs pretty fast, at least when I was trying it again.

1 Like
6

I am guessing that they are sending those jobs over to a larger cluster node now, or larger nodes are simply available now.

These “which node” tuning details are modified constantly in response to the runtime metrics from everyone’s jobs and the cluster/nodes that are available. Threading is added and other details. The jobs all run inside of a container. This means a job can be run “anywhere” and produce the same result but the speed will depend on how robust a node is (memory and related details).

If this interests you, these are the “general” rules for the cluster resources that the UseGalaxy servers use as a baseline. A search with the tool name will find all the related lines. The default right now is 2 cores but a server administrator could decide to increase this more (but probably not less!) if they have the resources available.