LEfSe LDA plot issue

77 views

Skip to first unread message

Akriti Singh

unread,

Sep 19, 2019, 1:01:24 PM9/19/19

to LEfSe-users

Hello,

My LEfSe LDA plot has missing taxa... How can I fix this? Please see the plot below and a sample of the text file I used. I used the Greengenes database in QIIME2-2018.8 (gg-13-8-99-515-806-nb-classifier_2018.qza).

Thanks,

Akriti

LEfSe_Factor1_Species L_18 Sept 2019.png

ID	Singh-2800	Singh-2801	Singh-2838	Singh-2847
Factor1	lowmed	high	lowmed	lowmed
k__Archaea\|p__Euryarchaeota\|c__Methanobacteria\|o__Methanobacteriales\|f__Methanobacteriaceae\|g__Methanobrevibacter\|s__	0	0	0.000282	0
k__Bacteria\|__\|__\|__\|__\|__\|__	0	0	0	0
k__Bacteria\|p__\|c__\|o__\|f__\|g__\|s__	0	0	0	0
k__Bacteria\|p__Actinobacteria\|c__Actinobacteria\|__\|__\|__\|__	0	0	0	0
k__Bacteria\|p__Actinobacteria\|c__Actinobacteria\|o__Actinomycetales\|__\|__\|__	0.000172	0	0	0
k__Bacteria\|p__Actinobacteria\|c__Actinobacteria\|o__Actinomycetales\|f__Actinomycetaceae\|g__Actinomyces\|__	0	0	0.000343	0

Jérémie Auger

unread,

Sep 19, 2019, 3:59:36 PM9/19/19

to LEfSe-users

What do you mean by "missing taxa"? The lines that only read " __ " ? Those are in fact only unattributed to this taxonomic level (like you have "k__Bacteria|__|__|__|__|__|__" in your input file).

In those cases, (better when you have less in the dataset, like only 1-2 "__" ! but still..) you can download (or generate if working locally! ) the "differential features" directory. The names of the files in there tell you what higher taxa the "__" come from.