Identifying DEG between development stages
1
0
Entering edit mode
cicely.day • 0
@e356d75d
Last seen 2.0 years ago
France

Hi there, I am trying to analyse a data set to identify differentially expressed genes between plant development stages. Currently, I have generated the data for the wildtype. I have followed the manual for deseq using the htseqinput pipeline, and generated the dds dataset. The problem I am having is comparing the expression between the 8 developmentally stages I have. I do not have "treated" versus "untreated" as is commonly the case in the online forums I have read.

I believe I need to use the LTR test. However, my design has just 1 variable (stage) and so I am confused as to how to write a formula for my reduced model. I don't want to compare the DEG in a pairwise manor against my reference, but overall.

Can anyone give me some advice?

Thanks in advance, C

sampleFiles <- grep("readCount",list.files(directory),value=TRUE)
sampleStage <- c('12dpaWT', '12dpaWT', '2mmWT','2mmWT','2mmWT', '4DPAWT', '4DPAWT','4DPAWT', '4mmWT', '4mmWT', '4mmWT', '8dpaWT', '8dpaWT', '8dpaWT','8mmWT', '8mmWT','8mmWT','AntFleurWT','AntFleurWT','AntFleurWT','AntFruitWT', 'AntFruitWT', 'AntFruitWT', 'FRUITO')
sampleTable <- data.frame(sampleName = sampleFiles,
                          fileName = sampleFiles,
                          Stage = sampleStage)
sampleTable$Stage <- factor(sampleTable$Stage)

library("DESeq2")
ddsHTSeq <- DESeqDataSetFromHTSeqCount(sampleTable = sampleTable,
                                       directory = directory,
                                       design= ~ Stage)
ddsHTSeq # Provides user with table containing information on the matrix

keep <- rowSums(counts(ddsHTSeq)) >= 10 # Remove rows which have less than 10 reads, rowSums calculates the sum of rows of a matrix or an array, counts is the number of occurences)
dds <- ddsHTSeq[keep,] # This removed ~10 000 elements, reducing size of matrix

dds$Stage <- relevel(dds$Stage, ref = "FRUITO")
reduced <-  #Need to create a reduced formula ?? 
dds <- DESeq(dds, test="LRT", reduced= )
plant Transcriptomics DESeq2 r • 1.1k views
ADD COMMENT
0
Entering edit mode
Basti ▴ 780
@7d45153c
Last seen 7 days ago
France

You were almost there, just have a look at the DESeq2 vignette part "Likelihood ratio test" : http://bioconductor.org/packages/devel/bioc/vignettes/DESeq2/inst/doc/DESeq2.html

dds <- DESeq(dds, test="LRT", reduced=~1)
ADD COMMENT
0
Entering edit mode

Hi Basti, thank you for that quick response. Yes I had read that, but maybe I have interrupted it wrong... as I have just 1 design variable, would I simply write my reduced formula as:

dds <- DESeq(dds, test="LRT", reduced=~1)

I was trying to create a new variable, but was unsure which design term to remove as i have just one, i.e. here:

dsHTSeq <- DESeqDataSetFromHTSeqCount(sampleTable = sampleTable,
                                       directory = directory,
                                       design= ~ Stage)

Cheers!

ADD REPLY
1
Entering edit mode

I don't understand why would you want to create a new variable, if you want to test for multiple stages at once as you formerly described, just keep the DESeq() function like this

ADD REPLY

Login before adding your answer.

Traffic: 611 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6