# Understanding beta-group-significance permanova pairwise statistics **URL:** \ **Category:** General Discussion\ **Created:** [June 24, 2019, 8:01pm UTC](https://forum.qiime2.org/t/understanding-beta-group-significance-permanova-pairwise-statistics/10396 "2019-06-24T20:01:28Z")\ **Posts on this page:** 4\ **Page:** 1
**Author:** ![kcelona](https://forum.qiime2.org/letter_avatar_proxy/v4/letter/k/e9c0ed/32.png) [@kcelona](https://forum.qiime2.org/u/kcelona)\ **Post date:** [June 24, 2019, 8:01pm UTC](https://forum.qiime2.org/t/understanding-beta-group-significance-permanova-pairwise-statistics/10396/1 "2019-06-24T20:01:28Z")
Hi, I have looked through the forum to the best of my ability to find answers to my questions, but have come up short on a few questions. I realize they are probably simple answers, but Qiime2 is new to me, and so is the statistics involved. To quickly summarize my work, I have a collection of serum samples from aerosol challenged rhesus macaques that I tested with a serodiagnostic assay to look for immune responses. The serum samples can be categorized by collection date, such as pre-challenge (-7), day 7, day 14, and day 28. I used the qiime diversity beta/diversity PcoA (Thanks to Matt Dillon's help!) to create a principle coordinate analysis of my samples in hopes that the samples would stratify by collection date due to the immune responses produced.... They did! Commands I ran are below: biom convert -i table.txt -o table.biom --to-hdf5 qiime tools import --type 'FeatureTable[Frequency]' --input-path table.biom --output-path table.qza qiime diversity beta --i-table table.qza --p-metric braycurtis --o-distance-matrix dm\_bray\_curtis.qza qiime diversity pcoa --i-distance-matrix dm\_bray\_curtis.qza --o-pcoa pcoa\_bray\_curtis.qza qiime emperor plot --i-pcoa pcoa\_bray\_curtis.qza --m-metadata-file metadata.tsv --o-visualization emperor\_bray\_curtis.qzv # Open up `emperor_bray_curtis.qzv` at [https://view.qiime2.org](https://view.qiime2.org) I took this a step further and explored the qiime diversity beta-group-significance plugin to see if the difference between my "collection day" groups were significant. I chose a PERMANOVA method, pairwise comparison, and 999 permuations (no particular reason why). Commands I ran below: qiime diversity beta-group-significance --i-distance-matrix --m-metadata-file --m-metadata-column --p-method --p-pairwise --p-permutations --o-visualization Thanks to previous forums, the graphs produced make sense to me. Here is an example where I have distances 1-7 (these are early terminal draws, essentially another group) ![image](https://forum-qiime2-org.s3.dualstack.us-west-2.amazonaws.com/original/2X/d/d110e804a5dbda31588186bc4e81599417f38e1f.png) I'd like to ask how to interpret the statistics that are associated with this analysis. I understand that the pseudo-F value is difficult to interpret and due to many variables. If there is more information I can learn about how to interpret the pseudo-F, I'd be delighted to know! I'd like to know more about the p-value, if it is safe to say a p-value less than 0.05 is deemed significant, and also how to understand the q-value. ![image](https://forum-qiime2-org.s3.dualstack.us-west-2.amazonaws.com/original/2X/3/3747fff2a30e4895f16e6f90d904de62a1171efb.png) If you need more information, please let me know! ![igg_bc_1-7](https://forum-qiime2-org.s3.dualstack.us-west-2.amazonaws.com/original/2X/2/27e2de6092dab109d2f20f7963b363c1bc0d33a6.jpeg) ![image](https://forum-qiime2-org.s3.dualstack.us-west-2.amazonaws.com/original/2X/3/3747fff2a30e4895f16e6f90d904de62a1171efb.png) ---
**Author:** ![Nicholas\_Bokulich](https://forum.qiime2.org/user_avatar/forum.qiime2.org/nicholas_bokulich/32/19937_2.png) [@Nicholas\_Bokulich](https://forum.qiime2.org/u/Nicholas_Bokulich)\ **Post date:** [June 25, 2019, 11:09am UTC](https://forum.qiime2.org/t/understanding-beta-group-significance-permanova-pairwise-statistics/10396/2 "2019-06-25T11:09:31Z")
> [@kcelona](#): > > If there is more information I can learn about how to interpret the pseudo-F, I’d be delighted to know! You have probably read through these, but the following are the best resources for info on interpreting PERMANOVA: [http://cc.oulu.fi/~jarioksa/softhelp/vegan/html/adonis.html](http://cc.oulu.fi/~jarioksa/softhelp/vegan/html/adonis.html) [https://mb3is.megx.net/gustame/hypothesis-tests/manova/npmanova](https://mb3is.megx.net/gustame/hypothesis-tests/manova/npmanova) > **[Permutational analysis of variance](https://en.wikipedia.org/wiki/Permutational_analysis_of_variance)** > > Permutational multivariate analysis of variance (PERMANOVA), is a non-parametric multivariate statistical permutation test. PERMANOVA is used to compare groups of objects and test the null hypothesis that the centroids and dispersion of the groups as defined by measure space are equivalent for all groups. A rejection of the null hypothesis means that either the centroid and/or the spread of the objects is different between the groups. Hence the test is based on the prior calculation of the dis In... and of course the original publication (linked from those links) > [@kcelona](#): > > if it is safe to say a p-value less than 0.05 is deemed significant Yes (or rather the q value if you are doing multiple group comparisons) > [@kcelona](#): > > how to understand the q-value The q-value is the p-value corrected for multiple tests. See here: > **[False discovery rate](https://en.wikipedia.org/wiki/False_discovery_rate)** > > In statistics, the false discovery rate (FDR) is a method of conceptualizing the rate of type I errors in null hypothesis testing when conducting multiple comparisons. FDR-controlling procedures are designed to control the FDR, which is the expected proportion of "discoveries" (rejected null hypotheses) that are false (incorrect rejections of the null). Equivalently, the FDR is the expected ratio of the number of false positive classifications (false discoveries) to the total number of positiv Th... Since you are doing a longitudinal analysis you may also want to check out q2-longitudinal, which would allow you to do paired tests and various longitudinal analyses that are appropriate for your data. See here for a tutorial: [https://docs.qiime2.org/2019.4/tutorials/longitudinal/](https://docs.qiime2.org/2019.4/tutorials/longitudinal/) Good luck! ---
**Author:** ![kcelona](https://forum.qiime2.org/letter_avatar_proxy/v4/letter/k/e9c0ed/32.png) [@kcelona](https://forum.qiime2.org/u/kcelona)\ **Post date:** [June 26, 2019, 6:27pm UTC](https://forum.qiime2.org/t/understanding-beta-group-significance-permanova-pairwise-statistics/10396/3 "2019-06-26T18:27:52Z")
Thank-you! This is great! ---
**Author:** ![system](https://forum-qiime2-org.s3.dualstack.us-west-2.amazonaws.com/original/3X/2/1/21af5fe23cb6f4579467c66a9ed94e55274ca7bd.svg) [@system](https://forum.qiime2.org/u/system)\ **Post date:** [July 28, 2019, 12:29am UTC](https://forum.qiime2.org/t/understanding-beta-group-significance-permanova-pairwise-statistics/10396/4 "2019-07-28T00:29:10Z")
This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.