Question

Simulate data using DESeq2

0

Entering edit mode

boymin2015 • 0

@boymin2015-8514

Last seen 9.3 years ago

Japan

I have successfully generated the simulation data using the function " makeExampleDESeqDataSet" of DESeq2,

But I have no idea to distinguish which genes are DEG and which are not.

DESeq2 makeExampleDESeqDataSet • 2.6k views

ADD COMMENT • link updated 9.3 years ago by Michael Love 42k • written 9.3 years ago by boymin2015 • 0

score 0 · Answer 1 · 2015-07-31

0

Entering edit mode

andrew.j.skelton73 ▴ 370

@andrewjskelton73-7074

Last seen 8 months ago

United Kingdom

have you followed the vignette?

ADD COMMENT • link 9.3 years ago andrew.j.skelton73 ▴ 370

0

Entering edit mode

Yes, I even check the R code. The following is my data simulation code：

deseq2 <- makeExampleDESeqDataSet(n = 10000, m = 2 * 5, betaSD = 4, interceptMean = 4, interceptSD = 2, sizeFactors = rep, dispMeanRel = function(x) 4/x + 0.1)

Can U tell me the information about DEG and non-DEG in the deseq2?

ADD REPLY • link 9.3 years ago boymin2015 • 0

0

Entering edit mode

Well you can run

dds <- DESeq(deseq2)
res <- results(dds)

but, by generating that dataset, you're at the equivalent stage of 1.3 in the users guide.

ADD REPLY • link 9.3 years ago andrew.j.skelton73 ▴ 370

score 0 · Answer 2 · 2015-07-31

0

Entering edit mode

Michael Love 42k

@mikelove

Last seen 11 hours ago

United States

You should always first check the help file for the function by typing a "?" and the function name:

?makeExampleDESeqDataSet

...
Value:

a ‘DESeqDataSet’ with true dispersion, intercept and beta values in the metadata columns.  Note that the true betas are provided on the log2 scale.
...

So you have:

mcols(dds)$trueBeta

Note that this simulation is a very simple one, and the parameters are not based on any dataset. Better would be to estimate the mean and dispersion pairs from a dataset (as we did in the DESeq2 paper) and feed these vectors into the function (the input arguments can be vectors).

ADD COMMENT • link 9.3 years ago Michael Love 42k

0

Entering edit mode

Thanks for the reply. I also got the $trueBeta. But the problem is how can I define DEGs or non-DEGs since it is just a vector of numbers. Usually, I thought the users can obtain the information about the DEG or non-DEGs of simulation data before evaluating the DE analysis methods.