Hello everyone! I hope that you are enjoying the best of health.
I analysed my 16S data using QIIME2 pipeline. Although I have completed the data analysis, I could not generate publication quality figures using QIIME2 plugins. Then I used the .biom files generated through QIIME2 and ran phyloseq to make publication quality figures. The figures that I am getting with phyloseq are show sliughlty different results than the results shown by the figures genearted through QIIME2 plugins. I have the followintg questions:
Are there any QIIME2 plugins that can create publication quality figures?
Is it normal to get slightly different results, especially for alpha and beta diversity metrics, using phyloseq?
Does phyloseq just plot the data, or does it calculate metrics on its own before generating the visualizations?
There are many ways to make 'publication quality' figures! I like using R for print publications.
Note that phyloseq graphs are ggplot objects, so there are easy to modify to meet publication guidelines. The graphs in the .qzv files are easier to make but harder to modify, as you have learned! If you use Qiime2 thought the API, you also access the underlying data and make your own custom plots with seaborn or plotly or whatever.
Yes, because there's a unifrac beta-diversity bug in phyloseq:
Note that bug was found in 2018, and phyloseq has not been under active development since then.
Phyloseq can do both! So you could run unifrac with Qiime2 to get valid distances, then graph it with Phyloseq and ggplot2.
Doing both in phyloseq risks that bug:
ordu = ordinate(GP1, "PCoA", "unifrac", weighted=TRUE)
# ⚠️ phyloseq::ordinate(distance="unifrac") is bugged, see https://github.com/joey711/phyloseq/issues/956
