Core microbiome: at treatment- not sample-level

[tcar...@uncc.edu]

Dear all,

I have 16S rRNA data for 300 samples: 100 from site A, B, and C. I would like to determine the 'core microbiome' across sites. Instead of comparing at the sample-level (i.e., which OTUs are observed in each of my 300 samples), I wish to compare at the site-level (i.e., which OTUs are present at the 3 sites, even if they're not in every sample)  

The shared_phylotypes.py script determines the # of OTUs (at the site-level) for this inquiry, while compute_core_microbiome.py operates at the sample-level. If you, however, collapse all samples per site, you may subsequently generate the 'core microbiome' across sites. The issue with the latter is that you are left with 3 data points in your revised .biom (i.e., site A, B, and C) and cannot perform, for example, a PCoA on that 'core' community to test how this community differ between sites. 

Would anyone know a way around this?

Thanks, in advance, for the assistance. 

Regards,

Tyler

--

Tyler J. Carrier

  Ph.D. Student, Reitzel Laboratory

  Department of Biological Sciences

  University of North Carolina at Charlotte

  Charlotte, NC 28223

Fastest Communication

  e-mail: tcar...@uncc.edu

[TonyWalters]

Hello Tyler,

I think you should be able to generate beta diversity output for 3 samples, but only the 2d plots. E.g. run each of these scripts, in this order, on your collapsed OTU table:

http://qiime.org/scripts/single_rarefaction.html

http://qiime.org/scripts/beta_diversity.html

http://qiime.org/scripts/principal_coordinates.html

http://qiime.org/scripts/make_2d_plots.html

You might also consider comparing the beta diversity of the uncollapsed samples with the compare_categories.py script: http://qiime.org/scripts/compare_categories.html