Differential abundance analysis: Plugin error from gneiss

Hi all,

I am trying to run the command:
qiime gneiss ols-regression for the Multivariate response linear regression and I am getting this error message:
Plugin error from gneiss:

[('/tmp/qiime2-archive-xusbcxp1/dd9cde43-3230-48a7-a185-a625897d5e17/provenance/metadata.yaml', '/tmp/qiime2-provenance-1eeg1ybk/artifacts/dd9cde43-3230-48a7-a185-a625897d5e17/metadata.yaml', "[Errno 13] Permission denied: '/tmp/qiime2-provenance-1eeg1ybk/artifacts/dd9cde43-3230-48a7-a185-a625897d5e17/metadata.yaml'"), ('/tmp/qiime2-archive-xusbcxp1/dd9cde43-3230-48a7-a185-a625897d5e17/provenance/citations.bib', '/tmp/qiime2-provenance-1eeg1ybk/artifacts/dd9cde43-3230-48a7-a185-a625897d5e17/citations.bib', "[Errno 13] Permission denied: '/tmp/qiime2-provenance-1eeg1ybk/artifacts/dd9cde43-3230-48a7-a185-a625897d5e17/citations.bib'"), ('/tmp/qiime2-archive-xusbcxp1/dd9cde43-3230-48a7-a185-a625897d5e17/provenance/action/action.yaml', '/tmp/qiime2-provenance-1eeg1ybk/artifacts/dd9cde43-3230-48a7-a185-a625897d5e17/action/action.yaml', "[Errno 13] Permission denied: '/tmp/qiime2-provenance-1eeg1ybk/artifacts/dd9cde43-3230-48a7-a185-a625897d5e17/action/action.yaml'"), ('/tmp/qiime2-archive-xusbcxp1/dd9cde43-3230-48a7-a185-a625897d5e17/provenance/action', '/tmp/qiime2-provenance-1eeg1ybk/artifacts/dd9cde43-3230-48a7-a185-a625897d5e17/action', "[Errno 13] Permission denied: '/tmp/qiime2-provenance-1eeg1ybk/artifacts/dd9cde43-3230-48a7-a185-a625897d5e17/action'"), ('/tmp/qiime2-archive-xusbcxp1/dd9cde43-3230-48a7-a185-a625897d5e17/provenance/VERSION', '/tmp/qiime2-provenance-1eeg1ybk/artifacts/dd9cde43-3230-48a7-a185-a625897d5e17/VERSION', "[Errno 13] Permission denied: '/tmp/qiime2-provenance-1eeg1ybk/artifacts/dd9cde43-3230-48a7-a185-a625897d5e17/VERSION'"), ('/tmp/qiime2-archive-xusbcxp1/dd9cde43-3230-48a7-a185-a625897d5e17/provenance', '/tmp/qiime2-provenance-1eeg1ybk/artifacts/dd9cde43-3230-48a7-a185-a625897d5e17', "[Errno 13] Permission denied: '/tmp/qiime2-provenance-1eeg1ybk/artifacts/dd9cde43-3230-48a7-a185-a625897d5e17'")]

Debug info has been saved to /tmp/qiime2-q2cli-err-9y_g3vnd.log

I am seeing there some "Permission denied". Not sure what that means. I am using version 2018.8. I ran the metadata tabulate to look for errors or blanks and seems to be good. I am attaching it.
tabulated-sample-metadata.qzv (1.2 MB)

This is whole command I am trying to run:
qiime gneiss ols-regression --p-formula "Nat_SampID+Time+Diet+Date+Time_Point" --i-table balances.qza --i-tree hierarchy.qza --m-metadata-file mapping_file_resistant_starch_only65_differential_analysis_dec21.txt --o-visualization regression_summary.qzv

Thanks for the help!

Hi @idapantoja, not exactly sure what is going on here. Could you post the log file /tmp/qiime2-q2cli-err-9y_g3vnd.log?

It does look like this sort of error has appeared on the qiime2 forums before. Its possible that upgrading to the newest version of qiime2 will do the trick.

Hi @mortonjt,

Thanks so much for your quick response!

Attached the file:
error.txt (2.7 KB)


Ok, it looks like you have miniconda running under a root environment

Do you have your own user space? If so, maybe try installing the new version of qiime under your own personal account.

Hi @mortonjt,

I am using qiime2 in a linux server environment. I can’t update qiime2 by myself. That is the only option to be able to run this command?


Hi @idapantoja — who installed qiime2 on this server? It is a little unusual that it was installed as the root user — this is going to cause all kinds of problems with running tools in this environment (not just QIIME 2, but everything in this root miniconda). Maybe you should check in with your sysadmin and see what the story is…

1 Like

Agree with @thermokarst, its pretty weird if you don’t have full control over your environment.

Is it possible for you to install qiime locally? If you were able to run any of the qiime commands (i.e. saving files, moving files, …), you should be able to install miniconda to your local directory and then install qiime through conda.