Question: export edgeR normalised OTU table
0
0
Entering edit mode
Pozdrawiam • 0
@Pozdrawiam-24676
Last seen 11 months ago

Hello, I want to normalise my microbiome data (OTU table) using edgeR package and export my normalised matrix ( I want to test some normalisation methods, normalised data will serve to calculate beta diversity and visualise by PCoA). I went through bioconductor forum and biostars and I still don't know want is the proper way to export normalised data from edgeR.

So, I read my matrix and calcNorm:

d <- DGEList(counts = d, group=group)
d = calcNormFactors(d)


what happens next?:

1st option

d = estimateCommonDisp(d, verbose=TRUE)
normalised_df <- d$pseudo.alt  2nd option normalised_df <- cpm(d, normalized.lib.sizes = T)  3rd option dgList <- estimateCommonDisp(dgList) dgList <- estimateTagwiseDisp(dgList) normalised_df <- t(t(dgList$pseudo.counts)*(dgList$samples$norm.factors))


I have an impression that every post claims something different and at this point I am very confused. Thank you!

microbiome edgeR • 389 views
0
Entering edit mode

Is anything wrong with the answers you were provided here (https://www.biostars.org/p/487419/) citing the manual that 2nd option is preferred and third is discouraged? It is good practice to indicate crossposts by adding links to the other community.

0
Entering edit mode

I just wanted to get a 3rd opinion, as your answer was different from the first one:)