Plugin error from diversity:

User Support
, , ,
1

I am having problem to generate alpha diversity. Until this step everything is ok, but when I do this command: qiime diversity alpha-group-significance
--i-alpha-diversity core-metrics-results/faith_pd_vector.qza
--m-metadata-file mapping_fileJ793.txt
--o-visualization core-metrics-results/faith-pd-group-significance.qzv

I receive this message:

Plugin error from diversity:

Metadata does not contain any columns that satisfy this visualizer's requirements. There must be at least one metadata column that contains categorical data, isn't empty, doesn't consist of unique values, and doesn't consist of exactly one value.

Debug info has been saved to /var/folders/ll/_2t9tfrd5rd5wx01ksb41gs00000gn/T/qiime2-q2cli-err-qnmq1cn3.log

I dont know what is happening with my metadata. I did the validation on https://keemei.qiime2.org/ and it said that was everything alright.

my metadata (.txt) is annexed below

03%20PM
03%20PM1428×636 83.3 KB

3

Hi @Giulia_Marchioro! One thing to keep in mind is that if your --i-alpha-diversity has fewer samples than your actual metadata file, only the metadata records that have matches will be used. This is important to consider in the context of rarefaction - depending on the even sampling depth you used, you might have removed many more samples than anticipated, which could explain the specific error message you are seeing:

How about you run the following and send the results our way? Then we can figure out where to go from there!

qiime metadata tabulate \
  --m-input-file core-metrics-results/faith_pd_vector.qza \
  --o-visualization core-metrics-results/faith_pd_vector.qzv

This will generate a table of alpha diversity values found in that Faith's PD vector. We are interested in which samples are present in this vector.

Keep us posted! :t_rex:

1 Like
5

I just run this command and this is my output

52%20PM
52%20PM1980×432 22.2 KB

So, because there is few samples that's why it happened right? (just to make sure). And the problem is not with my metadata but is because of the quantity of the samples being analyzed, right?

The sequence counts is very low with this primers pair and these sample.

23%20PM
23%20PM1260×812 24.9 KB

7

Ah ha! That is it! What was the rarefaction depth you specified when running core-metrics? The metadata that is actually being used are these two rows:

md

So that error message above makes sense! The available metadata is only comprised of columns with unique values or only one value.

1 Like
9

1066 because was the rest was too low!

10

Actually, I did with 1066 and 312, and for both was the same message.

But, it is happening because the sequence counts are too low, right? and I cant go forward?

11

Well, if you rarefied to 312, this should've worked --- maybe you mixed up files while you were testing that out?

No, not exactly --- it is happening because you chose a rarefaction depth that removed most of your samples (this can happen no matter how many reads are present in your data set). You can use a few tools to explore the impact of rarefaction on your data:

With that said, these do seem to be pretty low frequencies, you might also want to back up and revisit your denoising process, and see if you can tune your parameters there (for example, if you used q2-dada2 to get to this point, taking a closer look at your demux summarize quality plots and tightening your trim/trunc params).

Keep us posted! :t_rex:

12

Oh yeah!! It worked, I did something wrong for this 312 rarefaction.

Thank you for everything. Im gonna take a look why there is very low counts!! Thank you

1 Like
13

This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.