Pairwise differences according to taxonomy qiime2

Hey,
I'm trying to do pairwise acording to taxonomy of two time points,
this is the code:
qiime longitudinal pairwise-differences --m-metadata-file metadata_OK147na.tsv --i-table OK147na_table_filter_frequency_no_mitochondria_no_chloroplast_WG_AC_NA_out_relative-frequency.qza --p-metric featuretable --p-state-column weight_gain_POINT_A_C --p-group-column TreatmentNum --p-state-1 A --p-state-2 B --p-individual-id-column Patient --p-replicate-handling random --o-visualization OK147na_taxonomy_AB_pairwise-differences.qzv

and this is the error I got:
Plugin error from longitudinal:

metric must be a valid metadata or feature table column.

Debug info has been saved to /tmp/qiime2-q2cli-err-k1rabsar.log

I don't know how to fix it ..
Can I get some help please :slight_smile: :qiime2:

Welcome to the forum, @Tina_Khone!

Based on your error message, it looks like you are telling QIIME 2 to use a metric column that doesn’t exist in your Metadata. Have you looked over the longitudinal analysis tutorial yet? It has a good description of the command you are running, and the second blue “note” in it (near the top) has some helpful tools for checking what columns are available.

Let us know how it goes!

2 Likes

My metadata doesn’t have that column, isn’t the feature table the one to contain it?

I’m not sure how to approach this,

Hi!
You are providing

as a column name, but in fact you should indicate a name of the column with numerical data to analyse across your groups. It can be a certain taxa or ASV ID, since you provided a feature table as an input.

Hey!
I’m really sorry, but I just started working with Qiime2, so i’m not that good (yet) :laughing:
I don’t understand which file I need to do this kind of analysis …
I chose “relative frequency” (as a table) from a previous post on the subject.

In my metadata I have the sample ID, but not the taxonomy data,
how do I choose the correct file?

Thank you

Hey!

No problem, I remember myself struggling with a terminal after first installation of Qiime2 :grimacing:

So, if I am right, you would like to know if there are any differences in the abundances of certain taxa between some groups of samples. It is also why you provided a feature table as an input together with your metadata file. Let us know if in fact you want to perform another comparisons.

In the plugin description, it is written:

Note that "metric" can be derived from a feature table or metadata.

So, metric column could be in your metadata or feature table.
If you want to compare some taxonomy instead of ASV, you should use collapsed to desired taxonomy level feature table. If you want to compare ASV, than just provide feature-table as it is.
You can visualize a feature table to check all taxa names/ASVs, decide which taxa / ASV you want to compare, copy it as it is labeled in the table and past in your command as the metric.

1 Like

So if i’m using the collapsed qza, what is the matric? taxonomy?

Any taxa you want to compare between groups.

As I wrote earlier, you can:

If I understand correctly, I can’t compare all taxa in the two time points?

You are right, it will not compare all taxonomy, only specified one.

1 Like

Is there an analysis in Qiime2 that can do that? :sweat_smile:

Hi!
Maybe you should take a look on ANCOM, GNEISS or LEfSe (not a qiime2 plugin)

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