Qiime gneiss ols-regression

(Yogesh Gupta) #1

Hi,

I do have the multiple soil samples and I have given three treatment to these soil sample: T1, T2, T3. So I want to see the treatment effect on microbial abundance in these samples.

#[sample metadata columns are like this:

  sample_id  Tissue     Treatment    Villages              Region

    soil-01      soil       T1          Menjana            Rajshahi

    soil-02      soil       T2          Khola               Mymensing

could you please suggest me how I can design --p-formula for this analysis:

qiime gneiss ols-regression
–p-formula -------
–i-table balances.qza
–i-tree hierarchy.qza
–m-metadata-file sample-metadata.tsv \

0 Likes

(Jamie Morton) #2

Hi @Yogesh_Gupta, what have you tried so far?

It may not be a bad idea just to try to test for differences between your treatments (i.e. --p-formula Treatment ).

There are also a couple of previous threads on this (which can be found with the search bar functionality). See below

https://forum.qiime2.org/t/gneiss-use-of-combined-categories-in-covariates/7865/5

0 Likes

(Yogesh Gupta) #3

Hi @mortonjt,

thanks for your help.

I used the following commands:

qiime gneiss ols-regression  --p-formula Treatment --i-table balances.qza --i-tree hierarchy.qza --m-metadata-file metadata3.tsv --o-visualization regression_summary.qzv

and then for heat map:

qiime gneiss dendrogram-heatmap --i-table soil-c10-mins2-table.qza --i-tree hierarchy.qza --m-metadata-file metadata3.tsv --m-metadata-column Treatment --p-color-map seismic --o-visualization heatmap.qzv

I attached the heatmap, could you please help to understand this heat map, and which balance should i further investigate using the the below command:

qiime gneiss balance-taxonomy \
  --i-table table.qza \
  --i-tree hierarchy.qza \
  --i-taxonomy taxa.qza \
  --p-taxa-level 2 \
  --p-balance-name .....
  --m-metadata-file sample-metadata.tsv \
  --m-metadata-column Subject \
  --o-visualization y0_taxa_summary.qzv

Thanks a lot.
Yogesh

0 Likes

(Jamie Morton) #4

@Yogesh_Gupta - we need more information. What do the regression summaries look like?

0 Likes

(Yogesh Gupta) #5

Hi @mortonjt,

Thanks, I have attached the regression summary.

Kind Regards
Yogesh

0 Likes

(Jamie Morton) #6

ok … those summaries really don’t look good. Do you see any separation with beta diversity?

If not, I’d be inclined to say that there is no signal.

0 Likes

(Yogesh Gupta) #7

Hi @mortonjt,

Thanks for your reply.

Initially, I run all samples together: I do have 3 tissue: root, shoot and soil and three treatments: T1 , T2 ,T3.

For gneiss differential abundance, I only processed with soil samples with filtering the data, but earlier diversity analysis were performed all together.

should I filter the soil sample and perform the diversity analysis again? for gneiss analysis I also perform feature table filtering, should I do same filtering before doing diversity analysis.

Kind Regards
Yogesh

0 Likes