Question: DESeq2 multiple factor design

I have an experiment where I got 2 factor, the first one is the cell line, and the second one is the treatment.

In this, one cell line is derived from the other, the derived cell line received a inducible gene.

I then treated both cell lines with the inducer or vehicle.

I want to set the factors so DESeq2 compared the mother cell line +/- inducer to obtain the list of genes that are altered by the inducer alone, to the get the list of genes that are altered from the derived cell line +/- inducer and compare to the first result.

as in factors A, B, C and D, I want AxB=AB and CxD=CD, then compare ABxCD.

is that something DESeq2 can do?

I tried setting the factor levels like that:

dds$Cond <- factor(dds$Cond, levels = c("PRH_D","PRH_V","PC3_D","PC3_V"))

and got that:

##[1] "Intercept"           "Cond_PRH_V_vs_PRH_D" "Cond_PC3_D_vs_PRH_D"
##[4] "Cond_PC3_V_vs_PRH_D"

anyone have any tip on how i could set this levels?

keep in mid I am quite new to this.

thanks in advance.

rna-seq

• 2,236 views • link

updated 4.7 years ago by gaelgarcia05

• written 6.2 years ago by veteudmar •

Please use the formatting bar (especially the code option) to present your post better. You can use backticks for inline code (`text` becomes text), or select a chunk of text and use the highlighted button to format it as a code block. I've done it for you this time.
code_formatting

Also,

keep in mid [sic] I am quite new to this

is not how one should phrase that they're new to a field in a formal/professional setting. You can simply say "I am quite new to this" without instructions on what others should do with that information.

• 0 views • link

written 6.2 years ago by Ram

Wait, what? I get asking people to format their posts better, but how is "Keep in mind [...]" wrong or not appropriate? Don't be so pedantic.

• 0 views • link

written 4.7 years ago by gaelgarcia05

1 answer

Hi veteudmar,

To keep it simple, I would simply create a new variable that had two levels, AB and CD, and then compare these in the standard way for any 2 group comparison in DESeq2.

Kevin

• 0 views • link

written 6.2 years ago by Kevin Blighe

Hello Kevin, thank you for the help.

I tried what you said with this code.

    cts <- read.csv(file="Counts.csv",header=TRUE,sep=",",row.names = 1)

coldata <- read.csv(file="sample_info.csv",header=TRUE,sep=",")

head(cts,2)

coldata


library("DESeq2")
dds <- DESeqDataSetFromMatrix(countData = cts,
                              colData = coldata,
                              design = ~ Drug + Gene)

keep <- rowSums(counts(dds)) >= 10
dds <- dds[keep,]


dds$Drug <- factor(dds$Drug, levels = c("Doxy","Vehicle"))
dds$Gene <- factor(dds$Gene, levels = c("PRH","Empty"))




dds <- DESeq(dds)
res <- results(dds)
res


resultsNames(dds)

##[1] "Intercept"            "Drug_Vehicle_vs_Doxy" "Gene_Empty_vs_PRH"

the result in the intercept was still not what I am looking for. I need basically a 4 way comparison. I need to know that the genes changed with the inducer are due to the gene expression modification and to due to the drug itself.

• 0 views • link

written 6.1 years ago by veteudmar •