Vertebrate_Mammalian library not added with kraken build function about peptides HOT 8 OPEN

Hi there,

The library should be compatible, but I think you're missing an intermediate step. When you download using the kraken commands, it automatically runs the kraken2-build --add-to-library command for each fasta file. You'll see that in the library/added folder of your database it'll have .fna and .fna.masked files for each of the files that have been added to the library (they also have just a random name). So what you should do is move all of the files that you copied to the library to a different folder and run something like this: for i in fasta_files/* ; do kraken2-build --add-to-library $i --db kraken2_database ; done (note also that the files need to be unzipped for this to work).
After you've run the add to library command then you can build the database that will then include everything that you want.

Thanks,
Robyn

from peptides.

But the command only downloads GCF_000001405.39_GRCh38.p13_genomic.fna.gz no matter if I use -human TRUE or FALSE. So what I want to do is to use a library with mammals in Kraken2 to identify animal species in my wgs sample.

from peptides.

OK I think I should have been more specific here - I think it probably isn't working because python is expecting either True or False (with only the first letter capitalized). I think either removing this human flag entirely (default is False, which you want) or saying --human False or if you modify the script by changing line 42 (which is currently human = args.human) to instead be human = False then any of these should fix the problem.

Thanks,
Robyn

from peptides.

I used the command with only the frist letter capitalized but I still only get the GCF_000001405.39_GRCh38.p13_genomic.fna.gz
file and no other fasta or fna files.

from peptides.

Oh that is strange. Did you try the other options, of either not including the flag at all or setting it within the code file?

from peptides.

OK I think I have figured it out now - were you using the --complete True option? If so, then the problem was that the script wasn't accepting 'chromosome' as being complete also (I think maybe this is something that is different with the latest NCBI update). If you download the version of download_domain.py that I have just added then this should now be fixed. Let me know if you still have issues!

from peptides.

Yes that was the problem - Thank you very much :)

from peptides.

No problem! Just a note that with the vertebrate_mammalian you may want to also include the incomplete/scaffold genomes as only about half of them are complete (you can see in the assembly summary assembly_level column how many there are and if they correspond to species of interest for you or not).

from peptides.