Different results with LEfSe on between Galaxy and Conda

Hi everyone,

Why did I see the different results on between galaxy platform and local environment (mac, conda)?
During galaxy was down, I used Lefse with local conda environment in mac.
The same data set was computed with the same settings such as class row, subject row, and normalization value).
Then I found that the results of galaxy and local were different, especially the order of KEGG number in output like the attached plots (sorry that the png file generated from local may be obscure…)

The codes were as follows:

lefse_format_input.py abundance_filtered_ko_exPre.txt all_exPre.in -c 2 -u 1 -o 1000000

lefse_run.py all_exPre.in all_exPre.res

lefse_plot_res.py all_exPre.res all_exPre.png 

I would like to know the reason why such difference was happened.
It will be appreciated if you can help, suggest, or comment me.

Thank you,

Hi @Sarasa ,
This is due to the version difference of lefse conda v1.0.8 and the galaxy/github lefse v1.1.2. Please use the manual installation from github as python setup.py install if needed in future rather than conda version since conda lefse is outdated.


Hi @sagunmaharjann thank you so much for your reply.
Now I understood the current situation.

Then I come up with additional questions as follows:

  1. what kind of command line could I use to check the version of conda lefse? I would like to see it but I think the option --version did not work for any command of lefse.
  2. could you tell me how to install the latest version (I mean v1.1.2) lefse from github? Is it just like, for example,
python setup.py install lefse

Is the above correct? Or, can I use lefse just after executing python setup.py install?
Which webpage can I refer to install lefse manually from github?

Thank you for your help!