DESeq2 time course experiments for multigroups comparison (>2 groups)
1
0
Entering edit mode
@candicechudvm-10038
Last seen 5.9 years ago

Hi all,

I have RNA-seq data from three groups (rapid, slow, and control) at three timepoints (T1, T2, T3). Each group has three replicates except for the control group. After performing the time course experiments in DESeq2, I got over 1000 DEGs.

Now, my simple question is:

Are those DEGs (1) behaving in a condition-specific way (aka rapid is expressed differentially from slow since T1 + rapid and slow are both differ from control) or (2) those DEGs just behaving differently from the control but there's no different between rapid and slow groups.

Here is my script:

countdata=read.table("all.genes.txt", sep=" ", header=TRUE, row.names=1)
condition <- factor(c(rep("control",2), rep("rapid",3), rep("slow",3),
                      rep("control",2), rep("rapid",3), rep("slow",3),
                      rep("control",2), rep("rapid",3), rep("slow",3)))
timepoints <- factor(c(rep("t1",8), rep("t2",8), rep("t3",8)))
sampleTable <- data.frame(condition = as.factor(condition),
                          timepoints = as.factor(timepoints))
d.deseq<-DESeqDataSetFromMatrix(countData = countdata,
                                colData = sampleTable,
                                design = ~condition+timepoints+condition:timepoints)
ddsTC<-DESeq(d.deseq, test="LRT", reduced = ~ condition+timepoints)
ddsTC$condition <- relevel(ddsTC$condition, ref="control")
     resTC<-results(ddsTC, alpha = 0.05)
DESeq2 Time course RNA-seq • 2.0k views
ADD COMMENT
0
Entering edit mode
@mikelove
Last seen 10 hours ago
United States

"control" is already the reference level of condition thanks to alphabetical sorting, but if it were not you would need to relevel before calling DESeq().

These DEG should exhibit changes across time which are condition specific: either rapid or slow or both exhibit a different expression profile over time compared to control.

ADD COMMENT
0
Entering edit mode

Thanks! If I understand correctly, for those genes that have low p-value in the time course experiments, they must express differentially from the control but not necessarily express differentially between rapid and slow?

I have picked the top 9 genes with the lowest p-value and got this picture. I am pretty sure that they expressed differentially from the control but not very positive about whether the rapid and slow groups have distinct expression.

ADD REPLY
0
Entering edit mode

You are detecting genes where there are any differences among the three lines. So you will also find differences where the lines diverge between rapid and slow (these just may not occur as often, or the effect is smaller for those genes, so they are not in the top 9).

ADD REPLY

Login before adding your answer.

Traffic: 737 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