[R] Interpretation of log odds

Georg Otto georg.otto at tuebingen.mpg.de
Tue Feb 12 11:45:32 CET 2008


Dear Corinna,

please do post questions related to bioconductor packages directly to
the bioconductor mailing list. You will have a much higher chance to
get a helpful answer.

The B-statistic is explained best explained in the limma user guide
(chapter 10), which comes with the limma package or from

http://www.bioconductor.org/packages/release/bioc/html/limma.html

Hope that helps,

Georg


"Schmitt, Corinna" <Corinna.Schmitt at igb.fraunhofer.de> writes:

> Hallo,
>
>> fit12<-lmFit(qrg[,1:2])
>> t12<-toptable(fit12,adjust="fdr",number=15000,genelist=qrg$genes[,1])
>> t12
>             ID     logFC         t              P.Value                        adj.P.Val                        B
> 1560	orf6.2714  -5,95911144	-7,504537362	0,0000000000000616459272630    0,0000000000430961073320568	20,85141454
> 8689	SW23	    2,709344216	 3,41198098	0,0006449261297639210000000    0,0396758555030764000000000	-0,62704052
>
>
> The data example comes from one experiment, where I want to know if genes are differentially expressed. As I saw in the onlinehelp for toptable the value B is the log odds that the gene is differentially expressed. When I now look at the B value 20,85141454 it says that the gene orf6.2714 is in 20,85% differentially expressed. Is it right? But how should I interpret the second example SW23 with a negative B value?
>



More information about the R-help mailing list