Hello,
I'm currently facing a puzzling situation in my microbiome project and could use some advice. I conducted an analysis to compare alpha and beta diversity between two groups, but no significant differences were detected in these diversity indexes.
However, when I proceeded to run LEfSe analysis in R, I found that ten different taxa were identified as potential markers. This unexpected finding has left me wondering whether it's normal to observe such discrepancies between diversity indexes and LEfSe results.
Moreover, I'm also seeking guidance on the normalization processes to use in the LEfSe R code. Currently, I'm employing the following code:
lef_out<-run_lefse(psrelative, group = "group", norm = "CPM",
kw_cutoff = 0.05, lda_cutoff = 2)
However, I've noticed that when I utilize different normalization methods, I obtain varying results. Could anyone provide insights into which normalization processes are most appropriate for LEfSe analysis and how they might impact the outcomes?
I'd greatly appreciate any input or suggestions on this matter.
Thank you for your help!
