Fitting a model, Regression


First, I would like to tell you guys how glad I am that you guys monitor this page and answer our question in a timely manner. Now to my question. I am looking to perform a regression on changes over time in the abundance of Ectomycorrhizal fungi. I do not have multiple sampling events, as I replaced time with space. As such, I have four points in time (Years) and data for each year. I wanted to see if If there is a tool I can use to perform this regression here. I tried the longitudinal plug-in, but it appears that I need multiple sampling, over time.

Thank you for your help

1 Like

Hi @Fabs,
QIIME 2 does not have a method to perform simple linear regression with abundance of a microbial taxon, the reason being that that abundance information is usually relative abundance, which is inappropriate for linear regression (breaks various statistical assumptions). If you really want to do that analysis, you will need to run it in R.

Looks like you have been using gneiss, which would be an appropriate method for this question — stick with that approach!

Correct — most (but not all) methods in q2-longitudinal require repeated sampling, so would not be appropriate for this test.

Good luck!

1 Like

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