[BioC] Re: B statistic (was: affylmGUI exporting)

[James Wettenhall]

Hi Edmund,

> I started using the toptables for differentially expressed genes. I am
> confused about "B statistic"? I gather that the larger the B
> statistic, the more significant the gene expression. But what sort of
> statistical test is it?

Please change the subject line to include "B statistic", if that 
is the focus of your question.

The B statistic is the log odds of differential 
expression (also known as a lod score).  Bascially if 
it is greater than 0 then you have more than a 50/50 
chance that your gene is truly differentially expressed 
given its fitted M value (listed in the top-table)

The original B statistic is described in Lonnstedt and Speed's 
"Replicated Microarray Data" :
http://stat-www.berkeley.edu/users/terry/zarray/TechReport/Baypap4d.pdf
(In particular, see Appendix A.1)

It has been modified slightly by Gordon Smyth in the limma 
package.

Regards,
James