Help for limma DE interpretation.

Hi everybody.

Maybe my questions will sound stupid but sorry I am a beginner in this domain: I am not sure about the interpretation of DE analysis from limma.

My goal is to list the genes that are up and down regulated between 2 conditions (cond 1 and cond 2).

cont.matrix <- makeContrasts(cond1 - cond2,levels=design)

...

summa.fit <- decideTests(fit.cont)
summary(summa.fit)

The output is:

   cond1 - cond2
-1     5
0    322
1     18

1) Does it mean that there are 5 genes that are up-regulated in condition 2 compared to condition 1 ?

And:

    topTable(fit.cont,sort.by="p",n="2")
                 logFC   AveExpr         t      P.Value    adj.P.Val          B
Gene1  7.6103683  4.490033 22.122651 3.967110e-11 2.045458e-08 14.7706257
Gene2 -5.3795057  4.839369 21.026610 7.215021e-11 2.045458e-08 14.3659544

2) What is the logFC and how to interpret it according to the value ?

Thanks for your help.

limma rnaseq

• 4,006 views • link

updated 9.5 years ago by Devon Ryan

• written 9.5 years ago by Picasa

1 answer

Yes
The log2 fold-change, so 1 means 2x change, 2 means 4x change and so on. Your example values are very large changes (probably to/from no expression).

• 0 views • link

written 9.5 years ago by Devon Ryan

Thanks for your help.

So for instance, Gene 1 is up regulated under condition 1 ? and Gene 2 is up regulared in COnditon 2 40x more than condition 1 ?

• 0 views • link

written 9.5 years ago by Picasa

Yes, or more completely, "gene 1 is up-regulated in condition 1 compared to condition 2".

• 0 views • link

written 9.5 years ago by Devon Ryan