Question

R script for getting Common DEG genes from all comparisons by limma analysis

0

Entering edit mode

rkp • 0

@rkp-10175

Last seen 6.9 years ago

Dear All,

I want to find the probe id of Common genes that are differentially expressed between all the comparisons by limma analysis. Any suggestions will be greatly appreciated.

With kind regards,

Rajesh

limma bioconductor Microarray R • 979 views

ADD COMMENT • link updated 8.0 years ago by Gordon Smyth 52k • written 8.0 years ago by rkp • 0

score 0 · Answer 1 · 2017-01-11

0

Entering edit mode

Gordon Smyth 52k

@gordon-smyth

Last seen 13 hours ago

WEHI, Melbourne, Australia

I assume you have already run something like this:

fit2 <- contrasts.fit(fit, contrast.matrix)
fit2 <- eBayes(fit2)

To find the genes that are DE for all contrasts:

results <- decideTests(fit2)
i <- rowSums(results != 0) == ncol(fit2)
fit2$genes[i,]

ADD COMMENT • link 8.0 years ago Gordon Smyth 52k

0

Entering edit mode

Dear Gordon,

Thank you very much for providing script but I have not found DEG for all contrast. It is showing NULL

> results <- decideTests(fit2)
> i <- rowSums(results != 0) == ncol(fit2)
> fit2$genes[i,]
NULL

ADD REPLY • link 8.0 years ago rkp • 0

0

Entering edit mode

Presumably you haven't assigned anything to fit2$genes. In a routine analysis, this should be a data frame containing annotation information for each row, e.g., probe IDs, gene symbols. See, for example, Section 4.8 of the limma user's guide. Otherwise, you could just use i to subset a vector of probe IDs corresponding to each row of fit2. For example, if you had previously defined the row names to be the probe IDs, then you could do rownames(fit2)[i].

ADD REPLY • link 8.0 years ago Aaron Lun ★ 28k