I have more questions about beta diversity and significance.... So by looking at the box plots, everything overlaps, which to me would indicate none of the sample days are significantly different from each other... But when looking at the table of the pairwise comparison table, all p values are <0.05, which would mean the comparisons are significant allrepssampledaypairwisebraycurtis.qzv (758.7 KB) . So I am very confused!
Also... what is the difference in p and q values??
Please read the following before posting!
Is this post a General Discussion topic? Have you reviewed the QIIME 2 Forum Glossary? Post to this category if you have a general question about microbiome science, bioinformatics, or other general questions, ideas, or topics to discuss. Please do not post questions here that have to do with technical support requests. Posts in this category are not guaranteed a response.
Boxplots represent the distribution of the distances. Your box is the interquartile region: 50% of your data falls within that range. (The rest of the distribution is represented by the tails.) While in some cases, you may see a big shift in your interquartile regions, you may not. Because of the size of your data and the spread, it may be harder to see that change. The p-values Im seeing somewhat accurately reflect what I'd expect looking at your boxplots. However, my unsolicited advice is that you may want to run a permdisp if you haven't already.
Here, a p-value is the raw p-value from doing the 999 permutations to get your test results. (The smallest value you can get here 0.001). The q-value has corrected this p-value for multiple hypothesis testing. Essentially, if we take the same distribution and resample it 100 times, we may see a signficiant difference at \alpha=0.05 5% of the time. So, when you run a bunch of tests you want to correct for that. I think uses FDR.