Greetings all, I am interested in using the gcrma method to analyze a set of data that extends over the mgu74av1 and mgu74av2 Affy chips. I suppose my best bet is to create a custom CDF environment and Probe package which contain only the common probesets across the two versions. Does this seem reasonable? I can think of two ways to approach this: 1) create custom files representing the Affy CDF and probe files (this I believe I can do). Then use tools in Bioconductor to create the R packages. I guess those include makecdfenv (with which I am not familiar) and whatever is needed to create the probe data. 2) work with R to extract the common data from the existing packages into new ones. I would welcome advice on which would be easier plus some pointers on what I need to do in R to achieve either one. If I am successful would such a combined mgu74av1 and mgu74av2 CDF environment be of sufficient interest to others to make generally available? Many thanks, Alex Stoddard

There is such a CDF package for mgu74a and mgu74av2 at http://www.stat.berkeley.edu/~bolstad/mixtureCDF/MixtureCDF.html Ben On Wed, 2004-06-30 at 11:55, Alexander Stoddard wrote: > Greetings all, > > I am interested in using the gcrma method to analyze a set of data that > extends over the mgu74av1 and mgu74av2 Affy chips. > > I suppose my best bet is to create a custom CDF environment and Probe > package which contain only the common probesets across the two versions. > Does this seem reasonable? > > I can think of two ways to approach this: > 1) create custom files representing the Affy CDF and probe files (this I > believe I can do). Then use tools in Bioconductor to create the R > packages. I guess those include makecdfenv (with which I am not > familiar) and whatever is needed to create the probe data. > > 2) work with R to extract the common data from the existing packages > into new ones. > > I would welcome advice on which would be easier plus some pointers on > what I need to do in R to achieve either one. > > If I am successful would such a combined mgu74av1 and mgu74av2 CDF > environment be of sufficient interest to others to make > generally available? > > Many thanks, > > Alex Stoddard > > _______________________________________________ > Bioconductor mailing list > Bioconductor@stat.math.ethz.ch > https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor -- Ben Bolstad <bolstad@stat.berkeley.edu> http://www.stat.berkeley.edu/~bolstad