I’m currently using MaASLin2, and I’m using it to find associations between 4 parulid warblers and the insect prey in their diet. I’ve been thinking of creating a visual for it with a heatmap a bit like this-
The snag I’ve been experiencing is that the coefficient value for each species with reference to the other species is a little different. So if I need to put in a effect size value for say Bird 1 for the insect 1 Leucauge venusta, its coeff in reference to Bird 2 is 4.8, but its coeff in reference to Bird 3 is 3.6. Is there any way I can get a single value for Bird 1 in reference to both Bird 2 and 3?
Hi,
We have some information on interaction effects in the MaAsLin 3 tutorial which might be helpful.
I’m not sure that I quite follow for your case; are you saying you have 9 groups, one for each cell in that 3x3? You’ll always need a comparison to at least one reference group since the coefficients and significance values from a linear regression tell you the difference between a particular group and the reference. Is it possible to select one of those groups as the reference in MaAsLin 2/3 and then report all coefficients relative to that reference group?
Will
I see. That might not work out for my planned visualization, but I can definitely use the presence of a reference group for other analyses, thanks!
