ADONIS: correct interpretation of interaction among factors

[ja.morillo (Jose A. Morillo)]

Dear Qiime2 Experts:

This is a question related to the interpretation of the output of the plugin "qiime diversity adonis", that I am starting to use. We have rRNA 16S V4 data from a simple experimental design. In a soil translocation experiment we moved soils from location L, H and M with a factorial design (so, samples from for ewample L were moved to L, H and M, and so on...). Then there is two factors 1) "Home" (the origin of soils), and 2) "Away" (the destination of the soils). Then the soils were left for a while to check what happened to the microbial communities.

I have run the analysis in qiime2-2019.4 with no troubles and produced the unw-Unifrac distance matrix. After exploring with PCOAs it seems clear that the "home" factor is very determinant (=samples tend to cluster by the origin), and the "away" needs further testing. In order to check this I run the following two comands:

qiime diversity adonis
--i-distance-matrix unweighted_unifrac_distance_matrix.qza
--m-metadata-file ../metadata.tsv
--p-formula Away:Home
--p-permutations 999
--o-visualization beta-significance/ADONIS-only_interaction--un-unifrac-sign.qzv

qiime diversity adonis
--i-distance-matrix unweighted_unifrac_distance_matrix.qza
--m-metadata-file ../metadata.tsv
--p-formula Away*Home
--p-permutations 999
--o-visualization beta-significance/ADONIS-2factors--un-unifrac-sign.qzv

I attach the two qzv and the metadata ADONIS-2factors--un-unifrac-sign.qzv (264.7 KB) ADONIS-only_interaction--un-unifrac-sign.qzv (264.5 KB) metadata.txt (846 Bytes) . Note that the comparison is not even because some samples we gone during the experiment but all home/away combinations have >3 samples.

I have problems to interpret the results. If the formula used for the model is Away:Home I get the result of the interaction among the two factors, that is clearly significant (p=0.001). When using the formula Away*Home, I get the results of each factor (I understand, treated independently) AND their interaction, that with this formula is clarly NOT significant (p=0.418). The df of the interaction also changed.

So: 1) something wrong?, 2) which is the correct way to test the interaction, 3) I used the order Away(: ,)Home, instead Home(: ,)Away because we measure what happend in the soils after being translocated, and want to link that to the origin to detect possible effects; is this correct?

Thanks a lot!

[Nicholas_Bokulich]

You should always test the main effects, not just the interactions (which can be difficult to interpret independent of the main effects).

It is normal for significance to change when taking other effects into account, adding interactions, and changing the order of the formula.

You may want to check out these other topics for more discussion about these actions:

Adonis vs. Anosim vs. Permanova

Hi! I'm a bit of a newbie with stats and I'm still trying to fully grasp things. I'm a bit confused on what exactly the test statistics are telling me for PERMANOVA (pseudo-F), ADONIS (R2) and ANOSIM (R), and what exactly I should be using for my question, that is: are there significant differences between locations? From what I understand of permanova, pseudo-F tells me the ratio of between-group variation and within-group variation. A larger pseudo-F value means that the between-group variat…

adonis, betadiver, and betadisp

An earlier post by @CarlyRae first introduced me to the fact that when examining beta diversity I need to consider not only the results of a multi or univariate anova, but also examining if the within-group distances to a centroid were equivalent among my groups. I'm wondering how to best interpret the situation where everything turns out to be significant. That is, if the results of both PERMANOVA and PERMDISP suggest significant differences, where do most people go next? I've run these tests b…