Relative abundance output - problem

Hello everyone,

I tried to obtain the taxonomy relative abundance, using the commands suggested in Processing Questions (visualizations and taxonomy assignment) (anyway, I will report the commands below).
I got this table (level-2-relativa.qzv) and I don’t understand the results, especially in the frequency column (in the picture below). Maybe, I missed something in the command.
image

I also attach the tables that I used.
Level-2-relativa.qza (34.0 KB)
tableVCOLL.qza (29.6 KB)
tableV.qza (184.4 KB)
taxonomyV.qza (198.1 KB)
level-2-relativa.qzv (324.4 KB)

qiime taxa collapse
--i-table tableV.qza
--i-taxonomy taxonomyV.qza
--p-level 2
--o-collapsed-table tableVCOLL.qza &

qiime feature-table relative-frequency
--i-table tableVCOLL.qza
--o-relative-frequency-table Level-2-relativa.qza &

qiime feature-table summarize
--i-table ./carlotta/Level-2-relativa.qza
--m-sample-metadata-file metadataV1V2.tsv
--o-visualization ./carlotta/level-2-relativa.qzv &

Thank you in advance for your help!!!
Best regards.

Carlotta

1 Like

Hey @charlotte!

Looks like this is our bad. Our visualization seems to kind of assume it’s dealing with counts everywhere, but it’s type definitely accepts RelativeFrequency and PresenceAbsence.

I’ve created an issue here to fix this.

In the meanwhile, you can download the .csv file at the bottom of the first tab in your visualization. However, these are the summed per-sample proportions which don’t have a super intuitive interpretation (and you are going to be fighting floating-point rounding errors).

What questions are you trying to answer with this data?

1 Like

Hi @ebolyen, thank you for your help.

I downloaded the file.csv, anyway my doubt was: for example, if I consider sample of “Bacteria_Proteobacteria; Frequency 22; # of Sample Observed In 25”, is 22 Frequency value the relative abundance of Proteobacteria? If not, what is it?

May you please explain to me what do you mean with “it’s type definitely accepts RelativeFrequency and PresenceAbsence”?

Thank you again for your support.
Best regards.

Charlotte

1 Like

Hi @charlotte!

Almost, but not quite. It’s the sum of the relative abundance of Proteobacteria of each sample. Imagine the following:

     Proteobacteria
S1  0.91
S2  0.83
S3  0.92
.
.
.
S24 0.43
S25 0.83

Total: 22.891869805385593
That number isn’t really all that useful.

The command qiime feature-table summarize accepts as input the following type:

FeatureTable[Frequency | PresenceAbsence | RelativeFrequency]

However, the visualization itself is probably better suited to only showing

FeatureTable[Frequency]

which would have prevented you from passing a

FeatureTable[RelativeFrequency]

and getting that relatively nonsense feature-sum of 22.891869805385593.

What is your current goal with looking at the level-2 taxonomy abundance? There’s probably a better way to get that information (usually qiime taxa barplot).

Thank you @ebolyen For your answer!!
I already did the QIIME taxa barplot, but the output is absolute abundances. My aim is to transform the absolute abundance table in relative abundance table, without doing it by my self in excel.
Do you know another way to do relative abundance table output?

Thanks again for your support.
Best regards.

Charlotte

Hey @charlotte,

Good news is you have already succeeded in that effort! It is just your table summary that is nonsensical. If you export Level-2-relativa.qza you will find a relative abundance .biom table which you can convert as needed via biom convert.

Let me know if you run into issues!

1 Like

Thank you @ebolyen for your help!!! I got it!!

Best regards.

Charlotte

2 Likes

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