FW: LEfSe output: number of features in the output highly exceed the number in the input data
Levesque, Nicole M.
Good morning, see below – can someone please reach out to Safaa?
Thanks,
Nicole
From: Safaa Wasof [mailto:Safaa...@ugent.be]
Sent: Tuesday, July 30, 2019 9:33 AM
To: Huttenhower, Curtis <chut...@hsph.harvard.edu>; Levesque, Nicole M. <leve...@hsph.harvard.edu>
Subject: Fwd: LEfSe output: number of features in the output highly exceed the number in the input data
Dear,
I was using LEfSe to compare ASV abundances between four different treatment classes. I agglomerate the ASVs at the family level before importing it to LEfSe. The input file contains 262 families. However, the output file contains 446 features (please find attached). I am wondering why that happens?
Does LEfSe look at higher level (in my case would be to the Class and Order levels) as well?
What is very confusing also is that some of the results seem just replicated several times with slightly different feature names. For ex. I have the feature called "Fungi.Ascomycota.Sordariomycetes.Unassigned.Unassigned", but then I got replicated results for Fungi.Ascomycota.Sordariomycetes.Unassigned AND Fungi.Ascomycota.Sordariomycetes.Unassigned.unassigned. I have no feature called "Fungi.Ascomycota.Sordariomycetes.Unassigned".
That is really very confusing and any help will be very appreciated.
Best regards,
Safaa
