Correlations in Qiime2

I’m new in the microbiome and bioinformatics field. Could anyone help me?
Can I use qiime2 to study the correlation between a specific phylum and a set of sleep duration values of a specific population?

Hi @Faiga,

Welcome to the field! In order for the community to be able to help with your questions you’ll have to be a bit more specific with your questions. Start with what kind of data do you have and from what platform does the data derive from. Are you starting with raw sequences from a machine or are you playing with pre-existing data and you’re only looking to analyse it?
Have you had a chance to check out the qiime2 tutorials and examples of some common workflows? They are very thorough and can walk you through right from the beginning even if you have no prior experience. Once you familiarize yourself with some of the basic concepts and goals then the community can provide more specific help and guidelines!
Good luck!

1 Like

Hi @Mehrbod_Estaki,
Thank you so much for trying to help. I started with fastaq files generated from my samples and a metadata that I built with all the information from each sample. I went over all the tutorials and I was able to work my way through the commands and generate the qza and qzv files.
My research deals with a group of people with a specific disease and has 2 phases, before and after an intervention. I measure specific parameters for the disease (before and after the intervention). At the moment I have data only from the first phase. I wanted to see if I could find any correlations between bacteria and these parameters…
I hope you can help.

1 Like

QIIME2 does not have a pearson/spearman correlation method for correlating features with metadata — the reason being that this is not really an appropriate test for compositional (e.g., relative abundance) data.

You may want to check out q2-gneiss. It is a complicated but powerful method that could be used to approach this question (correlate metadata with all features).

When you have your post-intervention data, check out q2-longitudinal — some of the tools in there are designed for pre/post studies.

I hope that helps!

1 Like

Thank you very much!


How can I use qiime2 to calculate the correlation between observation abundances and continuous-valued metadata?
How should I write the commands?
Any help will be appreciated.

Hi Faiga,
Were you able to get the help you needed through this post? If not, can you clarify your question so we can better help you?

Hi @sejsong,
Thanks for the reply.
Regarding the post, it doesn’t really help me. I would like to know if qiime2 calculates Pearson correlation coefficients for bacteria abundances and numeric parameters in my metadata file. For example, if the abundance of a specific bacteria is correlated to sleep duration, in minutes?

Hi @Faiga - please see @Nicholas_Bokulich’s post above.

As @Nicholas_Bokulich suggested, we recommend you take a look at q2-gneiss to get at the root of your questions. If you have any questions about q2-gneiss please just let us know here on the Forum!

Thanks! :t_rex:

2 off-topic replies have been split into a new topic: Are gneiss and q2-longitudinal correlation or regression methods?

Please keep replies on-topic in the future.

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