Moving pictures tutorial- qiime2 2021-4


Apologies for the super basic/OCD questions....

I am working my way though all of your wonderful tutorials. I am currently on the moving picture tutorial and have literally been copy and pasting the provided code. I am using the figures you guys have in the tutorial as my "answer key" say. I am getting slight differences which worries me that I am missing some big step.

I believe this may come from the import, but I am not sure.

I am able to successfully import the metadata, barcodes and sequences using wget, but when I "peek" at my sequences I get the following:

UUID:               3a5c21b5-061a-4f88-805c-ca9536af168c
Type:               EMPSingleEndSequences
Data format:        EMPSingleEndDirFmt

This may be a little OCD, but I am concerned why i am not getting the UUID of 23c4e11c-5579-48d7-9210-fad277a47004 that is provided in the tutorial.

I think this affected my later results as my emperor plots looked similar, but had definite differences. They weren't major (I meant to keep them, I will provide them if that will further assist you all) but I know subtle differences/errors on my end can be damning down the line.

Would this importing variation maybe the cause of this? Did I miss something major?

I would be happy to fully provide my code and/or screenshots of my plots, I just didn't want to clog up the forum more than I already have!

Thanks so much for your time and expertise!


No need to be concerned, but great catch!

In regards to the slight variation you are seeing with Emperor, that is due to differences in subsampling during rarefaction and is to be expected :slightly_smiling_face: For a more in-depth look at this, you might want to check out this video on our Youtube channel.

Returning to the UUID, a new one is generated with each run and should be unique. This helps keep exactly which data was used for each step straight and is also used in provenance tracking.

Hope this helps clear things up!


OKay! I just got worried because there were even differences in my box and whisker plots for subsequent analyses. Would this be due to difference in subsampling as well?

Will definitely take a look at that video! Thanks for your help!


Yep! Rarefaction produces a new, unique dataset that is a subset of the original data. Because these unique subsets are generated(with the intention) of being representative of the entire dataset, any analysis run on them should be more or less equivalent to others but will differ slightly.