Chapter 17 LEfSe

For the final part of this section, we will see if there are any statistically significant differences between the two sample groups. There are several ways in which this can be achieved but we will carry out LEfSe again.

Task: Go back to your shotgun_meta terminal (or create a new one and use . useshotgun). Then change directory to 4-FunctionalProfiling)

Thankfully we already formatted the file to work with LEfSe when we formatted it for hclust2

#LEfSe format
lefse_format_input.py \
diet_unstratified.relab.comp.tsv \
diet_unstratified.relab.comp.lefse \
-c 2 -u 1 -o 1000000
#Run LEfSe
lefse_run.py \
diet_unstratified.relab.comp.lefse \
diet_unstratified.relab.comp.lefse.out
#Produce LEfSe plot
lefse_plot_res.py \
--dpi 200 \
--format png \
#Below options added to increase max length of feature string
--max_feature_len 200 \
diet_unstratified.relab.comp.lefse.out \
diet_unstratified.relab.comp.lefse.png
#View plot
firefox diet_unstratified.relab.comp.lefse.png

Look at the output and see what pathways count as biomarkers for the 2 groups.