Alpha and Beta diversity between two samples

Hi!
I am trying to comprare the differences and similarities between two samples. I have just one replicates of each. I performed the following:

qiime diversity core-metrics-phylogenetic
--i-phylogeny rooted-tree.qza
--i-table merged_table.qza
--p-sampling-depth 56191
--m-metadata-file metadata.txt
--output-dir core-metrics-results

and I chose the sampling depth of 56191 as my Feature counts were 56192 for the sample with less feature counts (I attached the feature-table).

I have two problems.
First, the outputs are empty, for example, the jaccard_emperor.qzv shows nothing at all.
Second, when I want to continue with alpha diversity I get the following error:

qiime diversity alpha-group-significance \

--i-alpha-diversity core-metrics-results/faith_pd_vector.qza
--m-metadata-file metadata.txt
--o-visualization core-metrics-results/faith-pd-group-significance.qzv

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 /tmp/qiime2-q2cli-err-5u2xhs_u.log

I also attach the metadata file. I found in a topic that I had few replicates or samples to compare.
But, is there any way to compare the changes in taxon frequencies between two samples?
Thanks!
Melisa
merged_table.qza (33.4 KB)
metadata.txt (59 Bytes)

Hi @Melisa_Olivelli,
It looks like your table has both of your samples in it.

However this part of the error: "doesn't consist of unique values" is why your data doesn't work in qiime diversity alpha-group-significance. Basically, we need to be able to calculate intra-variability and inter-variability and if your only have 1 sample in a group then intra-variability can not be calculated.

I am not exactly sure why your jaccard has no samples in it. Could you include your Jaccard emperor plot for me to look at?

1 Like

Yes! This is the jaccard...
jaccard_emperor.qzv (873.1 KB)
Thanks!

Hi @Melisa_Olivelli,
Thanks for sharing!

This is not something I have seen before but after looking more into it, this makes sense.

Basically emperor is trying to take your distance matrix that only has one value (how different sample 1 is to sample 2.) and trying to project it to a three dimensional space. It can not do this because there is really only one value to create the plot. So its basically centering it and saying "tada! I have made a plot" (If you save the png you can see that axis 1 and axis 2 have 0% variability and are directly on top of each other)

Usually distance matrices are very big and horrible to interpret but given that you have only 2 samples, looking at your distance matrix in excel might be the best way to investigate your distances!

Hope this helps!
:turtle:

Hi @cherman2!
Thanks for the explanation. It makes sense. Is there any other function, besides looking into the excel, that can give me some information about the differences in the abundances of each taxa between samples? Or just the taxa bar plots of each one?
Thanks!

Hi @Melisa_Olivelli,
Yeah I think that most of QIIME 2 statistical commands will require more samples, so I think looking at distances in excel and looking at the taxabar plots of your samples are probably the best way to investigate the differences in your samples. Typically I would recommend ANCOM for this but I don't think it will be able to investigate differential abundance between just 2 samples. You could also look at your alpha diversity metrics in excel too.

Sorry to not have more info
:turtle:

1 Like

Many thanks!
Cheers!

1 Like

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