Bump GTDBTk to v2.4.0#664
Conversation
|
|
Thank you very much @dialvarezs ! A very nice thing to come abck from weekend too 😎 I',m going to run a test on a HPC (as we can't test GTDB on GitHub due to the massive database file), and if that works I will merge this in :) |
|
OK took a long while to get it running on the HPC (it's a new one which I completely forgot about 😅 ), and had to manually download the GTDB database as it's an offline cluster, then insufficient memory, and a bunch of other things... IT's running now but waiting for a 256GB node to be allocated to me... |
|
Don't worry, and thanks for the review! I forgot to update the db parameter 😅. |
|
Oh nice! What amount of memory were you generally having to use to get it to run? Given the database is like 40gb larger than the other I wonder if we need to bump the minimum amount |
|
Good question. I'm using the default 128GB, and until now I have seen no failed jobs. Now, looking into it, Slurm reports a MaxRSS of around 10GB, while Nextflow reports about 300GB. I think this is because pplacer uses RAM as disk cache, and switches to scratch space if memory is not enough. |
|
OK cool, thanks for that! I've discovered a bug ion the nf-core module (because our test data MAGs doesn't get hits in the GTDB database funnily enough), but once I've fixed that I'll update the module in your PR, run my test again, and then I thin kwe are pretty much good to go :) |
jfy133
left a comment
There was a problem hiding this comment.
LGTM And works for me too, thank you @dialvarezs ! Impactful first contribution 💪
|
Whelp I forgot to push the fixed module 😆 |
Update GTDBTk module to latest version, and update
summary_gtdbtk.pyto hande column name changes.PR checklist
nf-core lint).nextflow run . -profile test,docker --outdir <OUTDIR>).nextflow run . -profile debug,test,docker --outdir <OUTDIR>).docs/usage.mdis updated.docs/output.mdis updated.CHANGELOG.mdis updated.README.mdis updated (including new tool citations and authors/contributors).