LEfSe input file problem

Hi all,

  1. I’m trying to run the LEfSe galaxy, and I made 2 different input files.
    All the data are exactly same, except for taxonomy name format.


However, the results of LEfSe from 2 different input files are totally different.


I don’t know which make input files make totally different results. Please explain it.

  1. I have one more question about the LEfSe input file.
    When I see the input format from the LEfSe website, it is not perfectly matched with relative abundance files
    It looks like the upper taxonomy category includes all the relative abundance in the lower taxonomy class.
    I want to make the same format input files, so is there anyone to teach me some command line to make LEfSe input file?
1 Like

The first taxonomy format is preferred. See the tool form help for details (scroll down).

You don’t need to do data manipulation line-command. Review the Text/data manipulation tools in Galaxy. Many are based on line-command tools (and named after them) and include help if you are not familiar with them. Nearly any manipulation/reformatting is possible.



have you solved it?

I am not sure if these answers were satisfactory for the Jungmin’s problems. I have exactly the same problem.

  1. Lefse creates totally different results for the same input files with different formats.
  2. Input file format provided in Lefse website is not perfectly matched with relative abundances files. As Jungmin stated, in the website upper taxonomy category includes all the relative abundance in the lower taxonomy class. Is there a way to create such an input file from QIIME2 or is it possible to somehow change the format of the file provided by QIIME2?


Has anyone solved this?

Hi there,
I have a similar problem and can’t figure out what’s wrong. When I run B) LDA Effect Size (LEfSe) on data, I get this result:
“Number of significantly discriminative features: 0 ( 40 ) before internal wilcoxon
No features with significant differences between the two classes
Number of discriminative features with abs LDA score > 1.0 : 0”.
However, from previous analyses, I’m sure there are differences between my samples.
Screenshot of my data format below:

Many thanks,

Confirmed issue with the original tool authors. Might be fixed already. See the way microorganisms are named affect the final results of lefse - #6 by jennaj