search

biocore / qiime

Official QIIME 1 software repository. QIIME 2 (https://qiime2.org) has succeeded QIIME 1 as of January 2018.
GNU General Public License v2.0
286 stars 266 forks source link

pairwise ANOSIM in compare_categories.py #431

Open samikbagchi opened 12 years ago

samikbagchi commented 12 years ago

Dear QIIME team,

it'll be of great help, if you add pairwise ANOSIM in the next version of compare_categories.py

kind regards Samik

jairideout commented 12 years ago

We've had this request come up a couple of times on the QIIME forum (for PERMANOVA as well) and it doesn't look like R has this capability either, so it would be a nice addition.

elipowel commented 11 years ago

Also, with adonis please.

Perugolate commented 11 years ago

any idea how the pairwise anosims were done here? Convergence of gut microbiomes in myrmecophagous mammals http://onlinelibrary.wiley.com/doi/10.1111/mec.12501/abstract

jairideout commented 11 years ago

I'm not sure- maybe they used PRIMER to do this? Or perhaps they filtered their distance matrix to include each of the two groups they wanted to compare and ran ANOSIM on that? The paper and accompanying spreadsheet doesn't say whether/how they corrected the p-values for multiple comparisons. I recommend contacting the authors for details. If you find anything, it'd be great if you could post back here with details!

rob-knight commented 11 years ago

Jai is correct in that we ran the ANOSIMs using PRIMER. We ran a global test for differences among diet types, but PRIMER also provides pairwise comparison tests. The p-values we reported were not corrected for multiple comparisons. As noted by the PRIMER developers, we took care to interpret the pairwise analyses by examining the combination of the R statistic and the p-value, although we should note that the differences in our focal comparisons (myrmecophages vs other diet types) remained significant after correction.

Perugolate commented 11 years ago

Is there a problem with filtering the mapping file to retain the two groups of interest in order to do this with qiime?

sejsong commented 11 years ago

Filtering to groups of interest using compare_categories.py in qiime should do the same thing.

On Thu, Oct 17, 2013 at 1:32 PM, Perugolate notifications@github.comwrote:

Is there a problem with filtering the mapping file to retain the two groups of interest in order to do this with qiime?

— Reply to this email directly or view it on GitHubhttps://github.com/qiime/qiime/issues/431#issuecomment-26541799 .

jairideout commented 11 years ago

Great, thanks for the details @rob-knight @sejsong!