[BioC] RE: B statistic in limmaGUI

[Elizabeth Brooke-Powell]

Jason,

This is another point I have been trying to understand. In some experiments
I have had B values of over 22 and others only 6. Are these values directly
comparable or do they suggest that in a set where the numbers are low (say 6
and below) that most of the genes are unlikely to be differentially
expressed? Any thoughts welcome. 

I think this hits the point that looking at lists shouldn't be solely about
the numbers of genes you want to look at. If you see low values should you
look at look at a smaller top hits list (30 genes vs 50) and in a different
experiment where you have higher values you should look at more (top 100 vs
top 50). Any thoughts?

Liz

Molteno Building
Department of Pathology
University of Cambridge
Tennis Court Road
Cambridge, CB2 1QP
United Kingdom
 
Website: http://www.path.cam.ac.uk/~toxo/
Tel 01223 33 33 31(office) or 01223 33 33 29 (lab)

-----------------------------------------
Message: 20
Date: Thu, 19 Aug 2004 12:40:57 -0400
From: "Jason Hipp" <jhipp at wfubmc.edu>
Subject: RE: [BioC] B statistic in limmaGUI
Cc: <bioconductor at stat.math.ethz.ch>
Message-ID:
	
<1D7037B603266D4082E413CDB2FE857A08164FC5 at EXCHVS1.medctr.ad.wfubmc.edu>
	
Content-Type: text/plain;	charset="iso-8859-1"

What is the difference between a B value of 1 vs 2 vs 10?  How high can a B
value go?
Thanks,
Jason