MaAsLin2 normalization methods

Dear MaAsLin2 developers and users,

By default, the normalization method is set to TSS (total sum-scaling I assume).
But in case that my input data is species profile is already after TSS (i.e. relative abundances) - then I should change this parameter to normalization = “NONE” - correct?
I run MaAsLin2 on species relative abundance data, with and without TSS normalization, and got very different results, including both coefficients and qvals.

Furthermore, in case I am filtering my species table prior to MaAsLin2, for example, to remove rare or low prevalent taxa, then the sum of all species in a sample will be <100% now. In this case, should I add TSS normlaization or not when running MaAsLin2?

Thanks a lot

Hi @Vadim_Dubinsky - you are correct. Please use normalization = 'NONE' if the data is already TSS-normalized and filtered. Additionally, you can set min_abundance = -Inf to prevent further filtering. Hope this helps!