block desings (random effects) in PERMANOVAs

[jamorillo]

Hello, I am working in an experiment with several experimental blocks (different soil plots were we replicate the same treatments). I therefore need to include the term “block” in the PERMANOVA models (like a random effect, not like fix factor). Following the help command of adonis2 function in R, it seems the term “strata” is what I need. What about Microeco, is the term “group” in trans_beta$cal_manova passed to "strata" in adonis2? I tried but apparently the function trans_beta$cal_manova is ignoring whatever I specify in “group”. What is the best way to consider these blocks design in permanovas? Thanks!! jose

[jamorillo]

Hello again, I think I found the solution myself, just want to confirm this is not wrong. Apparently, I can pass adonis2 parameters to the function cal_manova (microeco). For example:

beta_Bray_16S$cal_manova( manova_set = "Site + N + P", by = "terms", strata = dataset_16S$sample_table$Block, permutations = 999 ) where: "beta_Bray_16S" is the distance matrix of the microeco object "dataset_16S" "Site + N + P" is the formula I want to test "by" and "strata" are parameters of vegan::adonis2 function and "dataset_16S$sample_table$Block" is the metadata column of my microeco object that contains the block factor.

Correct? Thanks, your suggestions are very appreciated. jose

[ChiLiubio]

Hi jose,

Yes. You are right. All the parameters not found by cal_manova function will pass to adonis2 function. Please use beta_Bray_16S$res_manova to check your result.

Best, Chi

[jamorillo]

Hello Chi, Thanks for implementing the function in this way. It is very flexible. Great. Best, jose