Hi all ,

We want only the Gene Expression for the experiment without replicate, So i just want to confirm that is this code is wright or not . I am providing code below.

And what should be the dispersion in "exactTest" . Please help me in that.

==============================================

library(edgeR)
Read_counts=read.delim("Table.txt",row.names = "Gene")
counts=subset(Read_counts,Read_counts$Control >= 10 | Read_counts$Case >=10)
y=DGEList(counts=counts,group = 1:2)
et=exactTest(y,dispersion = 0.04)
et
 
fdr=p.adjust(et$table$PValue)
cpmres <- cpm(y)[rownames(et),]
summary(de <- decideTestsDGE(et))
write.csv(et$table, file="DEresults.csv")
write.csv(fdr, file="fdrcorrection.csv")
raw=data.frame(et$table)
DEResult_Counts=merge(counts,raw,by="row.names",row.names=FALSE)
write.table(as.data.frame(DEResult_Counts),"DEResult_with_Count.tsv",row.names = FALSE)
detags <- rownames(y)[as.logical(de)]
plotSmear(et, de.tags=detags)
pdf("Volcano_Plot.pdf")
plot(raw$logFC,-log(raw$PValue,10),
     xlim = c(-6,6),ylim = c(0,9),
     col=ifelse(abs(raw$logFC)<=2 | -log(raw$PValue,10)<1 ,"black","red"),
     pch=".",cex=2,xlab="logFoldChange",ylab="-Log10(Pvalue)",main="Volcano Plot")   
abline(v=c(-2,2),h=1,lty=2)

legend(3,3,"Significant Differentially \n Expressed Genes",lty=1, col=c('red'), bty='n',
       cex=0.5)
text(-5,8,"Down Regulated Genes",cex=0.5)
text(5,8,"Up Regulated Genes",cex=0.5)
dev.off()

==========================================================================

You should probably run:

y <- calcNormFactors(y)

... before running exactTest. We can't really help you with the value of the dispersion, because you don't have any replicates from which to estimate it for your data set. For well-controlled experiments in the lab (e.g., with mice or cell lines), values below 0.05 are usually observed. For human patient data, this is often higher (0.1 to 0.5) and for single-cell data, it is higher still (>1). It really depends on what system you're working with, and how good your sample preparation is. That's the whole point of doing experiments with replicates.

Some other comments; your p.adjust call will use the default Holm correction to control the FWER, not the BH correction to control the FDR. You could fix this by setting method="BH", but it would be simpler to just use:

raw <- topTags(et, n=Inf, sort.by="none")$table

Incidentally, this means you can just cbind it with the counts to get DEResult_Counts, as the row order is the same.