Tools to assess intra-sample variability between treatments

General Discussion
1

I have an experiment with 8 replicates per sample and 12 treatments.
One thing I notice visually on the stacked barplot created in qiime2 is that one set of treatments have far more variability in microbiome composition compared to the other. It almost appears that one set of treatments stabilize the microbiome to a more defined community across all samples, while the other set shows a more chaotic nature between samples.

I don't think is is a technical artifact but rather something potentially biologically relevant. Curious what other users have used to assess and compare variability with microbiome data.

2

Hello,

This is a non-trivial question, but there is a publication that tackles it, please review: Temporal variability in quantitative human gut microbiome profiles and implications for clinical research | Nature Communications
You can treat your replicates similarly to temporal samples in this pub and they provide a variety of methods, which give answer to the variability question.

Cheers
V

3

In case you have not found it yet, here's the tutorial on longitudinal analysis in Qiime2

Here's how I would measure these things:

Lower beta diversity distance between subsequent samples

...to a more defined community across all samples

Lower beta dispersion after treatment

1 Like