Question

Deseq normalization with adjusting gene length

0

Entering edit mode

Maryam • 0

@maryam-17785

Last seen 5.4 years ago

Hi all,
I asked the question in Biostar and then I got the answer that I should use deseq2 normalization with gene length adjustment. Is it possible to use deseq2 normalization with gene length adjustment like as FPKM gene length adjustment? I have 250 samples from healthy and disease states and I want to integrate gene expression with metabolic model. Indeed, I need within sample normalization (adjusting gene length) and between sample normalization(Deseq2). Because FPKM is not a good normalization in cross-sample analysis I want to use deseq2 normalization but i should account gene length in the normalization too.

Any help appreciated

deseq2 normalization • 2.3k views

ADD COMMENT • link 5.4 years ago Maryam • 0

0

Entering edit mode

Here is the R code for whoever needs it.

cts <- as.matrix(read.csv("Count_data.txt", row.names = 1, header= TRUE, sep="\t"))
coldata <- read.csv("coldata.txt", row.names = 1, sep="\t")
dds <- DESeqDataSetFromMatrix(countData = cts, colData = coldata, design =  ~ condition)
txdb <- makeTxDbFromGFF("gencode.v19.annotation.gtf_withproteinids", format="gtf", circ_seqs=character())
ebg <- exonsBy(txdb, by="gene")
rowRanges(dds) = GRangesList(ebg)
FPKM = fpkm(dds)

ADD REPLY • link 5.4 years ago Maryam • 0

score 2 · Accepted Answer · 2018-12-12

2

Entering edit mode

Michael Love 41k

@mikelove

Last seen 18 hours ago

United States

We have an fpkm() function which does just this, but you need to provide the exons or the basepairs. See the function help page.

ADD COMMENT • link 5.4 years ago Michael Love 41k

0

Entering edit mode

Thanks Michael, but I do not want FPKM. I need deseq2 normalization which accounts gene length in normalization too. Is it possible?

ADD REPLY • link 5.4 years ago Maryam • 0

0

Entering edit mode

Again, this is what our function fpkm() does and you should read the help page for that function.