Question: Summary results interpretation in DeSeq2
0
gravatar for sharmila.ahmad
4 months ago by
sharmila.ahmad0 wrote:

Hi and good day,

Can anyone explain to me what should I do if got the following results after running DeSeq? What should I do with the outliers and low counts? I have had filtered the low count <5 before running the analysis.

### result for Sex (F as reference level)
res_Sex <-results(dds_full, name="Sex_M_vs_F") 
summary(res_Sex) 

out of 15895 with nonzero total read count
adjusted p-value < 0.1
LFC > 0 (up)       : 893, 5.6%
LFC < 0 (down)     : 1550, 9.8%
outliers [1]       : 19, 0.12%
low counts [2]     : 309, 1.9%
(mean count < 7)
[1] see 'cooksCutoff' argument of ?results
[2] see 'independentFiltering' argument of ?results

Thank you in advance.

Best, Mila

deseq2 • 91 views
ADD COMMENTlink modified 4 months ago by Michael Love24k • written 4 months ago by sharmila.ahmad0
Answer: Summary results interpretation in DeSeq2
0
gravatar for Michael Love
4 months ago by
Michael Love24k
United States
Michael Love24k wrote:

The outliers and low counts have been removed from the FDR bounded set, by setting padj to NA (see vignette). So you don’t have to do anything.

ADD COMMENTlink written 4 months ago by Michael Love24k

Thank you for your explanation.

Best, Mila

ADD REPLYlink written 4 months ago by sharmila.ahmad0

Hi Michael,

I have another few questions to ask;

1) regarding filtering the low gene counts and outliers using NOISeq before running the DESeq. It is necessary or recommended to do that? If yes, does it makes any difference since the DESeq itself will filter the outliers and low counts genes right?

2) I don't understand the principle of llfctreshold. As I understood by setting up the lfctreshold will make the stat/results more stringent and there is no a single value that can be used as a based or reference.

Again, thank you.

Best, Mila

ADD REPLYlink written 3 months ago by sharmila.ahmad0

It’s not necessary to filter before running DESeq2. You can filter out low count genes for efficiency, genes which for example have counts less than 10 for nearly all samples. But it’s not necessary. See vignette for details on the automatic filtering.

The paper and vignette describe the threshold in detail.

ADD REPLYlink written 3 months ago by Michael Love24k
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 16.09
Traffic: 333 users visited in the last hour