Hi All,
I just want to know if I understand which function to do.
If using compositional data (relative abundance) in Maaslin2... do we need still need to normalize using CSS/ TSS OR just transform (LOG or AST) the data....?
OR is it best to use count data then apply CSS AND LOG transformation?
Many thanks in advance,
J