Qiime and Stirrup Classifier for Vaginal 16S - classification and Assigning taxonomy

[Raghavee Venkatramanan]

Hi, 

I am trying to analyze samples for Vaginal microbiome. I started with the usual pipeline a where I pick_otus, prick_rep_set, assign_taxonomy and then make an otu table. I also using a tree and use it downstream. I used the Greengenes db as reference and used userach61 to assign taxonomy. I did not get classification at the species level, although my application requires me to do (Study about bacterial vaginosis). So I dug around to find that Vaginal Microbiome Consortium has both a classifier and a reference data in STIRRUPS. It classified almost 98% of my OTU sequences to a species successfully.

My problem is, I am not sure how to incorporate the results into the already existing assigned taxonomy text file that can be used to create a BIOM table. I am very familiar using BIOM tables for downstream application, so I am working to get it rather than try other means. 

The stirrup classifies only in the Genus and species level and nothing above. For the missing species level, I backfilled the values I got from STIRRUP classification in my existing assign taxonomy step. Most of it seems to be classified till the Genus level. Although there is almost 20 % of the otu seems to have "unidentified" as the taxa assignment. I have the genus and species classification from stirrup for it. I was wondering if I can just fill back g__;s__ values without the filling the above and still make it to work to build an OTU table ? As I am only interested in species classification I think I can afford to backfill these values. If not, can you please point me how such studies are performed ? 

Kindly let me know your thought and I appreciate your guidance. 

PS: Example classification results for the same OTU: 

Using QIIME and usearch-

denovo414714 k__Bacteria; p__Firmicutes; c__Bacilli; o__Lactobacillales; f__Lactobacillaceae; g__Lactobacillus; s__ 0.67 3

using stirrup : 

denovo414714|Lactobacillus jensenii|BT|1|100|89.8

As I can backfill my file in the above example, I will not be able to do so in the below example.

example: 

Using QIIME and usearch-

denovo234088    Unassigned      1.00    1

using stirrup: 

denovo234088|Lactobacillus iners|BT|1|100|80.6

Thanks, 

RV

[Jai Ram Rideout]

Hi RV,

If you're only really interested in the genus and/or species-level taxonomic assignments you received from Stirrup, you can format the taxonomic assignment file to use semicolon-separated assignments so that QIIME will understand it. For example:

denovo234088|Lactobacillus iners|BT|1|100|80.6

would become:

denovo234088<tab>Lactobacillus; iners<tab>BT<tab>1<tab>100<tab>80.6

Here, <tab> should be replaced with a tab character since the taxonomic assignment file is a tab-delimited file.

QIIME doesn't enforce a specific taxonomic nomenclature or formatting, it only requires the taxonomy assignment to be separated by semicolons.

Once you have this file, you can process it as usual (e.g. adding it to your OTU table) and perform taxonomic analyses such as summarize_taxa.py. If your taxonomic assignments only have two levels (genus, species), summarize_taxa.py will output a L1 table (genus) and L2 table (species). If you were to use a different taxonomy scheme, such as Greengenes, those levels would correspond to different taxonomic levels.

Hope this helps!

Jai