[BioC] GCRMA-induced correlations? LARGE Change in GCRMAexpression values

Wed May 14 23:02:20 CEST 2008

Hi,
I have just installed BioC 2.70 and haven't done any analyses yet to
compare with previous ones, but since GCRMA is very frequently used in
our facility I am curious to know this as well. So if Dr Wu could
enlighten us this would much be appreciated. Meanwhile i'll do some
comparions as well.

Regards,
Guido

BTW: we prefer to use the EB estimation for NSB (thus: gcrma(eset,
fast=FALSE), since the p-value distributions of two-group comparisons
look then as expected, in contrast to cases when the NSB was estimated
using the (default) MLE estimate (fast=TRUE). The expression estimates
for probesets differ also a lot between these modes. 
I am absolutely not an expert on this, but I have understood that
especially the MLE is 'susceptible' to the "correction" Lim et al
suggested. When applied to the (default) MLE, the p-value curves really
look nice (like the ones originally obtained using the EB mode), and
also the expression estimates are much more like those calculated using
the EB estimate. In other words, your observations *suggest* some
serious changes have been made to the GCRMA library, but I am sure Dr Wu
will inform us about this.  

> -----Original Message-----
> From: bioconductor-bounces at stat.math.ethz.ch 
> [mailto:bioconductor-bounces at stat.math.ethz.ch] On Behalf Of 
> Richard Friedman
> Sent: 08 May 2008 18:26
> To: Zhijin Wu
> Cc: bioconductor at stat.math.ethz.ch
> Subject: Re: [BioC] GCRMA-induced correlations? LARGE Change 
> in GCRMAexpression values
> 
> Dear Zhijin,
> 
> 	Has the change  that you described below been made in 
> or before GCRMA 2.12?
> 
> 	If so has the new GCRMA been benchmarked at does it 
> give comparable results to the old GCRMA?
> 
> 	I am noticing a large change in the absolute values of 
> intensity measurements
> 
> 	For the same probeset and and array normalized with the 
> same 8 other arrays done with GCRMA 2.10 I got 5.27 but for 
> GCRMA 2.12  I got 3.14
> 
> Does this sound like a change that can be expected between 
> versions 2.10 and 2.12
> 
> or does it sound as if I had made an error of some kind.
> 
> Thanks and best wishes,
> Rich
> ------------------------------------------------------------
> Richard A. Friedman, PhD
> Associate Research Scientist,
> Biomedical Informatics Shared Resource
> Herbert Irving Comprehensive Cancer Center (HICCC) Lecturer, 
> Department of Biomedical Informatics (DBMI) Educational 
> Coordinator, Center for Computational Biology and 
> Bioinformatics (C2B2)/ National Center for Multiscale 
> Analysis of Genomic Networks (MAGNet) Box 95, Room 130BB or 
> P&S 1-420C Columbia University Medical Center 630 W. 168th St.
> New York, NY 10032
> (212)305-6901 (5-6901) (voice)
> friedman at cancercenter.columbia.edu
> http://cancercenter.columbia.edu/~friedman/
> 
> In Memoriam,
> Arthur C. Clarke
> 
> 
> 
> On Feb 19, 2008, at 3:36 PM, Zhijin Wu wrote:
> 
> > Yes, to eliminate this artifact The truncated values will 
> no longer be 
> > adjusted in the next release of GCRMA.
> >
> > Jenny Drnevich wrote:
> >> Hi Zhijin,
> >>
> >> A client pointed out a July 2007 article by Lim et al. testing 
> >> different normalization/pre-processing methods for their 
> effects on 
> >> pairwise correlations between probesets (Bioinformatics 2007 
> >> 23(13):i282-i288; doi:10.1093/bioinformatics/btm201; full link 
> >> below). They reported that GCRMA introduced severe artificial 
> >> correlations between probesets; they looked for a cause and think 
> >> it's due truncation of low-intensity values after Non-Specific 
> >> Binding adjustment and then the Gene-Specific Binding 
> adjustment on 
> >> these truncated values. They also tested a specific 
> correction to the 
> >> GCRMA algorithm that appears to prevent the artificial correlation 
> >> and suggest that it become an option or even a default in the R 
> >> implementation of GCRMA.
> >>
> >> What do you think of this article? Are there any plans to 
> implement 
> >> their suggestion?
> >>
> >> Thanks,
> >> Jenny
> >>
> >> Comparative analysis of microarray normalization procedures:  
> >> effects on
> >> reverse engineering gene networks
> >>
> >> http://bioinformatics.oxfordjournals.org/cgi/content/full/23/13/
> >> i282? 
> >> 
> maxtoshow=&HITS=10&hits=10&RESULTFORMAT=1&andorexacttitle=and&andorex
> >> 
> acttitleabs=and&andorexactfulltext=and&searchid=1&FIRSTINDEX=0&sortsp
> >> ec=relevance&volume=23&firstpage=i282&resourcetype=HWCIT&eaf
> >>
> >>
> >>
> >> <http://bioinformatics.oxfordjournals.org/cgi/content/full/23/13/ 
> >> i282? 
> >> 
> maxtoshow=&HITS=10&hits=10&RESULTFORMAT=1&andorexacttitle=and&andorex 
> >> 
> acttitleabs=and&andorexactfulltext=and&searchid=1&FIRSTINDEX=0&sortsp 
> >> ec=relevance&volume=23&firstpage=i282&resourcetype=HWCIT&eaf>
> >>
> >> Jenny Drnevich, Ph.D.
> >>
> >> Functional Genomics Bioinformatics Specialist
> >> W.M. Keck Center for Comparative and Functional Genomics
> >> Roy J. Carver Biotechnology Center
> >> University of Illinois, Urbana-Champaign
> >>
> >> 330 ERML
> >> 1201 W. Gregory Dr.
> >> Urbana, IL 61801
> >> USA
> >>
> >> ph: 217-244-7355
> >> fax: 217-265-5066
> >> e-mail: drnevich at uiuc.edu
> >>
> >
> >
> > -- 
> > -------------------------------------------
> > Zhijin (Jean) Wu
> > Assistant Professor of Biostatistics
> > Brown University, Box G-S121
> > Providence, RI  02912
> >
> > Tel: 401 863 1230
> > Fax: 401 863 9182
> > http://stat.brown.edu/~zwu
> >
> > _______________________________________________
> > Bioconductor mailing list
> > Bioconductor at stat.math.ethz.ch
> > https://stat.ethz.ch/mailman/listinfo/bioconductor
> > Search the archives: http://news.gmane.org/ 
> > gmane.science.biology.informatics.conductor
> 
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: 
> http://news.gmane.org/gmane.science.biology.informatics.conductor
> 
> 