I am conducting a LEfSe analysis by R package “lefser” for 16S nucleotide sequences. But I got a result with nothing in table of result (“res”) :
This is analysis codes:
library(openxlsx)
library(“lefser”)
library(SummarizedExperiment)
exprs<-read.xlsx(“level-final_forLDA_R.xlsx”,rowNames=TRUE) # relative abundances
exprs<-as.matrix(exprs)
colData<-read.xlsx(“LDA_group.xlsx”,rowNames=TRUE) # sample information
colData<-DataFrame(colData)
data<-SummarizedExperiment(assays=list(exprs=exprs), colData=colData) # establish a SummarizedExperience object for analysis
data33<-data[,is.na(data$jun_pair)!=TRUE] # subset
res<-lefser(data33, groupCol = “timepoint”, checkAbundances=FALSE) # No1. LDA analysis
res<-lefser(data33, groupCol = “timepoint”) # No2. LDA analysis
both “res” show nothing.
level-final_forLDA_R.xlsx：
LDA_group.xlsx：
What did I do wrong? I appreciate any advice from you.
Best regards
Hi @SeeStar, apologies but while we are the original team behind LEfSe - it was written in Python and we maintain both the Python and galaxy versions of this implementation, although we are at the point of only applying patch fixes to this tool. We are not the team behind the lefser library. You can check out the Bioconductor site for assistance or report an issue with the library here.
I would also suggest checking out a more actively developed DA method tool like MaAsLin 2 or ANCOM-BC, papers like this one from Dr. Nearing do a great job of summarizing why you might want to favor newer methodologies.
Best,
Kelsey
Thank you for help and I will study this article.
Last week, I had to turn for Ubuntu 20.04 and installed lefse. Everything seemed to be going well, but I can’t use “lefse_run.py”.
I tried to solve this problem by “import rpy2.robjects as robjects” ( Halla install error - Downstream analysis and statistics / HAllA - The bioBakery help forum), but failed.
Can you help me? I didn’t know what to do next. Thanks.
Best regards
