Bioconductor: I have a question concerning normalization of data from Affymetrix Drosophila chip .cel files. This is a general question about how to normalize over multiple treatments. I have data from multiple (4) treatments -- actually a timecourse of 4 post-treatment intervals (0, 3, 6, and 24 hrs later) -- with 3-6 biologically independent arrays for each treatment. I also have 5 biologically independent control arrays -- the same 5 arrays serve as controls for comparison to each of the 4 time intervals. I have used RMA to compare each treatment separately against the controls to determine genes whose expression levels change significantly under each treatment. As a result, I normalized each of the four treatments to the controls separately -- that is I did four separate normalizations with the one set of controls and one time interval. Would it be more appropriate to normalize all of the the experimental arrays in the analysis to the controls at once rather than separately, one treatment at a time (as described above)? If so, is there a recommended way to do so? I attempted to normalize all my arrays at once in R but received an error message (which I unfortunately did not save). Many thanks. Paul Mack, Ph.D Department of Genetics University of Georgia Athens, GA USA 706-542-1578 (w) 706-542-3910 (fax) paulmack@arches.uga.edu

This question comes up at least monthly on this list. You can use the "searchable e-mail archive" to look at previous discussion. In summary, arrays that are to be analyzed together for differential expression or clustering should be normalized together. --Naomi At 09:01 AM 2/22/2005, Paul Mack wrote: >Bioconductor: > >I have a question concerning normalization of data from Affymetrix >Drosophila chip .cel files. This is a general question about how to >normalize over multiple treatments. I have data from multiple (4) >treatments -- actually a timecourse of 4 post-treatment intervals (0, 3, >6, and 24 hrs later) -- with 3-6 biologically independent arrays for each >treatment. I also have 5 biologically independent control arrays -- the >same 5 arrays serve as controls for comparison to each of the 4 time >intervals. I have used RMA to compare each treatment separately against >the controls to determine genes whose expression levels change >significantly under each treatment. As a result, I normalized each of the >four treatments to the controls separately -- that is I did four separate >normalizations with the one set of controls and one time interval. Would >it be more appropriate to normalize all of the the experimental arrays in >the analysis to the controls at once rather than separately, one treatment >at a time (as described above)? If so, is there a recommended way to do >so? I attempted to normalize all my arrays at once in R but received an >error message (which I unfortunately did not save). > >Many thanks. > > >Paul Mack, Ph.D >Department of Genetics >University of Georgia >Athens, GA >USA > >706-542-1578 (w) >706-542-3910 (fax) >paulmack@arches.uga.edu > >_______________________________________________ >Bioconductor mailing list >Bioconductor@stat.math.ethz.ch >https://stat.ethz.ch/mailman/listinfo/bioconductor Naomi S. Altman 814-865-3791 (voice) Associate Professor Bioinformatics Consulting Center Dept. of Statistics 814-863-7114 (fax) Penn State University 814-865-1348 (Statistics) University Park, PA 16802-2111

Paul Mack wrote: > Bioconductor: > > I have a question concerning normalization of data from Affymetrix > Drosophila chip .cel files. This is a general question about how to > normalize over multiple treatments. I have data from multiple (4) > treatments -- actually a timecourse of 4 post-treatment intervals (0, 3, > 6, and 24 hrs later) -- with 3-6 biologically independent arrays for > each treatment. I also have 5 biologically independent control arrays -- > the same 5 arrays serve as controls for comparison to each of the 4 time > intervals. I have used RMA to compare each treatment separately against > the controls to determine genes whose expression levels change > significantly under each treatment. As a result, I normalized each of > the four treatments to the controls separately -- that is I did four > separate normalizations with the one set of controls and one time > interval. Would it be more appropriate to normalize all of the the > experimental arrays in the analysis to the controls at once rather than > separately, one treatment at a time (as described above)? If so, is > there a recommended way to do so? I attempted to normalize all my arrays > at once in R but received an error message (which I unfortunately did > not save). In most cases it is preferable to normalize all of your samples together, and I don't see any compelling reason to do things separately here. I would bet that the error you received was something like 'Unable to allocate a vector of XXXX Mb', which implies that you ran out of memory. A reasonable workaround is to use justRMA(). Jim > > Many thanks. > > > Paul Mack, Ph.D > Department of Genetics > University of Georgia > Athens, GA > USA > > 706-542-1578 (w) > 706-542-3910 (fax) > paulmack@arches.uga.edu > > _______________________________________________ > Bioconductor mailing list > Bioconductor@stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor -- James W. MacDonald Affymetrix and cDNA Microarray Core University of Michigan Cancer Center 1500 E. Medical Center Drive 7410 CCGC Ann Arbor MI 48109 734-647-5623