Weird looking volcano plot
1
0
Entering edit mode
nico • 0
@80066fff
Last seen 14 months ago
Canada

Hello everyone,

I performed a DeSeq2 analysis and after I performed the lfcshrinkage command, I obtained very small log2fold changes for some of my results. As a consequence, my enhanced volcano plot looks very weird. For explaining my approach: I'm using TCGA-ovarian HT-Seq counts downloaded from the GDC Bioportal, Because there is a high amount of genes with 0s, I set the keep <- rowSums(counts(dds)) >= 1000 to a 1000 as seen and then continued with my analysis. After performing lfcshrinkage I obtained the following volcano plot.

Now I'm trying to figure out if something is wrong with my data or what is the reason for this weird looking volcano plot.

enter image description here

Has anybody obtained similar results like this?

Thank you very much for any help

DESeq2 volcano • 3.2k views
ADD COMMENT
0
Entering edit mode

You appear to be throwing out all genes with less than 1000 total reads. Is there a reason you're setting that threshold so high? I imagine this could introduce some peculiarities in your data set. You might reduce this threshold (the vignette sets it to 10!) and see how that changes your overall results.

ADD REPLY
0
Entering edit mode

Thank you for the quick reply. I set the default so high, to avoid the genes with so many 0s. That's the plot with the default setting (>10). enter image description here

ADD REPLY
0
Entering edit mode

It might be a good idea to explore your data to determine why you have so many zero count genes. Do these tend to occur in particular samples more often than others?

You could also try the alternative filtering method mentioned in the vignette where you set a requirement for each gene that you must have a certain number (X) of samples with read counts greater than a threshold:

Alternatively, a popular filter is to ensure at least X samples with a count of 10 or more, where X can be chosen as the sample size of the smallest group of samples:

keep <- rowSums(counts(dds) >= 10) >= X
dds <- dds[keep,]
ADD REPLY
0
Entering edit mode
@mikelove
Last seen 3 hours ago
United States

Can you post a plotMA plot? I'm guessing something is up with how you are labelling points.

ADD COMMENT
0
Entering edit mode

Hi Michael, I'm sorry for the late response.

Attached you will find first the plotMA plot before performing lfcshrinkage: enter image description here

And now the PlotMA plot after performing lfcshrinkage: enter image description here

Thank you very much for all your help!

ADD REPLY
0
Entering edit mode

This is a case where lfcShrink is more conservative than the null hypothesis testing. This can happen with few genes showing evidence of large LFC. For the volcano plot, maybe use type="normal" in lfcShrink or just stick to the MLE LFCs.

ADD REPLY

Login before adding your answer.

Traffic: 813 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