Question: how do compare with DESeq2
0
11 months ago by
lkianmehr0 wrote:

Hello to all,

I am starting to do DGE analysis with DESeq2 with a small example, as a first experiment I am going to compare two groups as example with DESeq2. I have two groups (ND and NF) and there 2 samples in each one (ND1 and ND2 in ND group, NF1 and NF2 in NF group). theses samples are COUNT file from HTseq-count. I am using this command in R for making groups, I am not sure they be correct;

gr <- factor(c(rep("ND", 2), rep("NF", 2)))
colData <- data.frame(group=gr, type="paired-end")
cds <- DESeqDataSetFromMatrix(cn2, colData, design= ~group)
cds <- DESeq(cds)
cnt <- log2(1+counts(cds, normalized=T))
dif <- data.frame(results(cds, c("group", "ND", "NF")))

I would be appreciate if would explain me they are right or not and how can I make group for DGE analysis

Leila

deseq2 • 220 views
modified 11 months ago by Michael Love24k • written 11 months ago by lkianmehr0
Answer: how do compare with DESeq2
0
11 months ago by
Michael Love24k
United States
Michael Love24k wrote:

This looks correct. Have you seen the DESeq2 workflow? This goes through the analysis at a slow pace to explain each step:

https://bioconductor.org/packages/release/workflows/vignettes/rnaseqGene/inst/doc/rnaseqGene.html

If you have further questions, feel free to post them.

Hi,

yes, I have. I made 5 groups to compare,  as below

gr <- factor(c(rep("WND", 12), rep("DND", 4), rep("WHF", 4), rep("DHF", 4))),

then I made contrast between each two group separately, like this:

dif <- data.frame(results(cds, contrast = c("group", "DHF", "DND")))

and tried to find up and downregulated Micro-RNA but I am surprised because there is not any significant case between them (log2FoldChange < -1 & padj < 0.05)!!!,

is needed to mention that I am amateur with using DESeq2, and I dont know exactly this method for comparing groups together is correct or not and how can I be certain about my results?

I would be appreciated if let me know your idea

You should be prepared to accept that there may not be any significant DEG between two groups. But you can also perform some checks, such as looking at the PCA plot to see if there is any expected groupings (conditions or batches, etc). What do you see in the PCA plot?