Question on Paired Design Implementation in MaAsLin3

Downstream analysis and statistics MaAsLin
1

Hi friends,

I have a paired-design study where each patient has two time points (before and after treatment). I’d like to test whether the microbiome community changes between these two time points are associated with BMI. Could you please advise on how to implement this analysis in MaAsLin3?

Thanks so much,
Ivy

2

Hi Ivy,

Assuming you have the BMI information before and after and everyone got the treatment, I would run the model ~ treatment*BMI + (1|subject) and use small_random_effects=TRUE. You’ll get 3 coefficients:

  1. treatment - the effect of treatment on the microbiome
  2. BMI - the association between BMI and the microbiome at baseline
  3. treatment:BMI - the change, relative to baseline, in the association between BMI and the microbiome after treatment

The small_random_effects are because you have 2 time points per person.

Let me know if that makes sense,

Will

3

Thank you so much!! @WillNickols