pick_closed_reference_otus.py - which assign_taxonomy_reference_seqs_fp is being used

[ecolog]

Hi,
I ran pick_closed_reference_otus.py and looking at the log got confused what reference_seqs_fp is being used. I passed the path for GG 13.5 taxonomy in my original script

pick_closed_reference_otus.py -i seqs_nonchimeric_denoised.fna -o otus -r /home/ecolog/Databases/gg_13_5_otus/rep_set/97_otus.fasta -a -O 6 --assign_taxonomy

However, the log file attached, doesn't explicitly seem to be using my  reference_seqs_fp,

parallel_assign_taxonomy_uclust.py -i otus/rep_set//seqs_nonchimeric_denoised_rep_set.fasta -o otus/uclust_assigned_taxonomy -T --jobs_to_start 6

instead the log present the system parameters from the qiime config and I see the reference_seqs_fp from GG 13.8

assign_taxonomy_reference_seqs_fp       /apps/compilers/python/2.7.9/lib/python2.7/site-packages/qiime_default_reference/gg_13_8_otus/rep_set/97_otus.fasta

Did I fail to assume that my reference taxonomy given in  pick_closed_reference_otus.py will be used for assigning the taxonomies?
If that is the case, should I rerun the process by changing my qiime config assign_taxonomy_reference_seqs_fp? and probably I need to update assign_taxonomy_id_to_taxonomy_fp from the qiime config too?

Thanks,
ecolog

[Daniel McDonald]

Hi ecolog,

The file provided to pick_closed_reference_otus.py with -r is used for the reference to compare the input sequences against. A taxonomy file can be passed using -t, which needs to have a 100% overlap in IDs with the reference sequences used, and which does not appear to be passed. However, it looks like your qiime config defaults have both the Greengenes 97% sequences and taxonomy setup, so things are being passed to assign_taxonomy.py as expected.

Best,

Daniel