[BioC] gcrma 1.0.0 vs 1.0.6

[Dick Beyer]

I was doing a comparison of earlier results done with gcrma 1.0.0 to some done with gcrma 1.0.6 and noticed some rather large differences.  Would someone look at the following code and see if the results should be comparable?

I did a 28 chip set of mgu74av2 chips with both versions of gcrma and did a ratio intensity plot of the expression value of the same chip.  I expected to see a straight line with zero slope, instead I got differential gene expressions up and down regulated up to 64 fold.  

For gcrma 1.0.6 I used:
mcatHeart.raw <- ReadAffy( 
filenames = rownames(pData(pd.heart)),phenoData=pd.heart)
mcatHeart.gcrma <- gcrma(mcatHeart.raw,type="fullmodel",fast=FALSE)

For gcrma 1.0.0 I used:
mcatHeart.raw <- ReadAffy( 
filenames = rownames(pData(pd.heart)),phenoData=pd.heart)
mcatHeart.gcrma <- gcrma(mcatHeart.raw)
 

In the gcrma 1.0.6 case I used:
base 1.9.0 
utils 1.9.0 
graphics 1.9.0 
stats 1.9.0 
methods 1.9.0 
affy 1.4.23 
Biobase 1.4.10 
tools 1.9.0 
gcrma 1.0.6 
annaffy 1.0.6 
KEGG 1.4.0 
GO 1.4.0 
reposTools 1.3.12 
mgu74av2 1.4.0 
multtest 1.3.5 
siggenes 1.0.5 
xtable 1.2-1 
mgu74av2cdf 1.4.3 

In the gcrma 1.0.0 case I used
base 1.9.0 
utils 1.9.0 
graphics 1.9.0 
stats 1.9.0 
methods 1.9.0 
affy 1.4.23 
Biobase 1.4.10 
tools 1.9.0 
gcrma 1.0.0
annaffy 1.0.6 
KEGG 1.4.0 
GO 1.4.0 
reposTools 1.3.12 
mgu74av2 1.4.0 
multtest 1.3.5 
siggenes 1.0.5 
xtable 1.2-1 
mgu74av2cdf 1.4.3 

I am also using a windows system.

Any help or suggestions would be greatly appreciated.  I am hoping to know which gcrma is correct.

Thanks very much,
Dick
*******************************************************************************
Richard P. Beyer, Ph.D.	University of Washington
Tel.:(206) 616 7378	Env. & Occ. Health Sci. , Box 354695
Fax: (206) 685 4696	4225 Roosevelt Way NE, # 100
			Seattle, WA 98105-6099
http://depts.washington.edu/ceeh/ServiceCores/FC5/FC5.html