q2-fragment-insertion - assigning taxonomy

[Jennifer_Bell]

I’m trying to follow this https://library.qiime2.org/plugins/q2-fragment-insertion/16/ tutorial and I’m currently stuck on this step

qiime fragment-insertion classify-otus-experimental
–i-representative-sequences rep-seqs.qza
–i-tree insertion-tree.qza
–i-reference-taxonomy taxonomy_gg99.qza
–o-classification taxonomy.qza

My problem is that the link to the taxonomy_gg99.qza is broken, nor is this reference taxonomy explained well. May I use the reference taxonomy I trained for my primers? Is this reference taxonomy 16S (515F/806R) specific? I’m guessing so since the previous parts of the workflow referenced the moving pictures tutorial, but it is not clear.

Finally, it appears that the file phylogeny.qza is not a standard .nwk tree file which I will need for further analysis (specifically calculating MNTD using the picante package). Is there was way to create this?

Thanks!

Thanks!

[Nicholas_Bokulich]

Welcome to the forum @Jennifer_Bell!

First, I am assuming you have seen this post and the other warnings about this experimental method:

q2-fragment insertion problems with Silva 128 taxonomy

Welcome to the forum @Rkubinski! Could you please provide the commands that you are running and the full error message (though that message seems clear enough, the full message may provide better context). I should add a word of warning: that method is experimental and does not really perform particularly well for taxonomy classification, so I would really recommend using the methods in q2-feature-classifier instead.

Second, it looks like that file was deleted so the tutorial is out of date (I have raised an issue to get this fixed). For now you might be able to access the old file here but I can’t guarantee if that will work (sorry! I am not a fragment-insertion developer but maybe @Stefan can help if that file does not work)

May I use the reference taxonomy I trained for my primers?

No — that trained classifier is specific to the qiime feature-classifier classify-sklearn method. It cannot be used with fragment-insertion.

Is this reference taxonomy 16S (515F/806R) specific? I’m guessing so since the previous parts of the workflow referenced the moving pictures tutorial, but it is not clear.

Yes, it uses Greengenes, which is 16S-specific

Finally, it appears that the file phylogeny.qza is not a standard .nwk tree file which I will need for further analysis (specifically calculating MNTD using the picante package). Is there was way to create this?

What happens if you use qiime tools export? The exported file should be .nwk

Good luck!

[Stefan (Stefan Janssen)]

Second, it looks like that file was deleted so the tutorial is out of date (I have raised an issue to get this fixed). For now you might be able to access the old file here but I can’t guarantee if that will work (sorry! I am not a fragment-insertion developer but maybe @Stefan can help if that file does not work)

@Nicholas_Bokulich is pointing to the correct file and is ready for use as is.

To give you some background: It basically is the original GreenGenes 13.8 file, located in the archive https://gg-sg-web.s3-us-west-2.amazonaws.com/downloads/greengenes_database/gg_13_5/gg_13_5_otus.tar.gz file taxonomy/99_otu_taxonomy.txt and was imported into a qza with the command qiime tools import --input-path 99_otu_taxonomy.txt --type "FeatureData[Taxonomy]" --output-path taxonomy_gg99.qza after adding the header line Feature ID Taxon to the file 99_otu_taxonomy.txt