Question

DESeq2 minimum p value

0

Entering edit mode

rw6 • 0

@rw6-12898

Last seen 7.0 years ago

When running differential expression analysis using DESeq2 the highest rank gene has a p value of 0. What is the lowest p value that can be reported by DESeq2?

deseq2 • 1.3k views

ADD COMMENT • link updated 7.0 years ago by Michael Love 41k • written 7.0 years ago by rw6 • 0

score 0 · Answer 1 · 2017-04-25

0

Entering edit mode

Michael Love 41k

@mikelove

Last seen 1 hour ago

United States

Like other parametric tests, there is not a minimum p-value.

> x <- 1; t.test(rnorm(10),rnorm(10,x))$p.value
[1] 0.04165338
> x <- 10; t.test(rnorm(10),rnorm(10,x))$p.value
[1] 4.523894e-15
> x <- 100; t.test(rnorm(10),rnorm(10,x))$p.value
[1] 1.836278e-33

ADD COMMENT • link 7.0 years ago Michael Love 41k

0

Entering edit mode

Thanks for the quick reply. I recognise that there is not actually a minimum p-value, but for one gene in my analysis the p-value is reported as 0. I assume that this is because there is some cut off where the number is too small to reported by the computer, as the value will not be exactly 0. Is it possible to know what value this p value must be lower than? (The lowest p value I have reported is 1.04E-188.)

ADD REPLY • link 7.0 years ago rw6 • 0

1

Entering edit mode

I actually don't know. If you're interested in ranking the genes, just use the Wald statistic. Once the null is certainly rejected, how small the p-value gets is not of great interest.

ADD REPLY • link 7.0 years ago Michael Love 41k