MaAsLin2 model for ko profile

Downstream analysis and statistics MaAsLin
1

Dear guys,

Our group met a problem when used MaAsLin2. We are trying to use MaAsLin2 to detect species and KOs which are respond to a additive. The program runs smoothly and I don’t get any error massage, as well as any signifcant results. We are wondering if there were indeed no bactera responded to this additive or we choosed inappropriate parameter, like the model.

There phenotype are considered: pen, weight and additive, the affect of pens need to be adjusted, so we put it as random_effects, weight and additive as fixed_effects, using default model.

fit_data = Maaslin2(input_data = pro,input_metadata = matedata,output = “ko/out”,fixed_effects = c(“weight”,“additive”), random_effects = c(“pen”),min_prevalence = 0.1,reference=‘additive,TZM’)

Do you think this model is suitable? or any suggestions for choosing a appropriate model?
Looking forward to your reply?

matedata.txt (674 Bytes) ko.profile.txt (2.9 MB)

2

@Panpan Based on the description, your model looks correct. If you have only tried the default model so far, I suggest trying out other models implemented in MaAsLin 2 (e.g. CPLM) although without any guarantee that it will give you top hits specially if the data has no signal. All the best!