How do I find taxa associated with continuous metadata variables?

General Discussion
1

Previously in QIIME I used “observation_metadata_correlation.py” to see if any of my species were associated with my continuous metadata variables.

In a post on a different topic (checkerboard) @Nicholas_Bokulich said “I think the reason is that the checkerboard score may not be appropriate for compositional data. That is the reason why none of the correlation methods used in observation_metadata_correlation.py were ported over).”

So my question is, how do I look for relationships with continuous data. It is my understanding that ANCOM and gneiss struggle with continuous data. My data is ITS so not suitable for anything requiring phylogenies

Any advice would be gratefully received.

1 Like
2

Hi @Cat,

This might be a good oppertunity for Songbird, which is designed for compositional data and should be able to handle a multivariate linear regression type model (including covariates!). I think ANCOM 2 in R also include multivariate models, so it’s possible?

Others may have more brilliance for you.

Best,
Justine

2 Likes