Mayte, I was skimming through some bioc-packages to get some inspiration and stumbled on this package: http://www.bioconductor.org/packages/2.10/bioc/html/virtualArray.html I've never used it, but it advertises itself to do exactly what you'd like, so perhaps you'll find it useful. -steve On Fri, Apr 27, 2012 at 5:01 PM, Steve Lianoglou <mailinglist.honeypot at="" gmail.com=""> wrote: > Hi, > > On Fri, Apr 27, 2012 at 4:51 PM, James W. MacDonald <jmacdon at="" uw.edu=""> wrote: >> Hi Mayte, >> >> The short answer is no. >> >> You could hypothetically create a mixture cdf that only contains probesets >> that are in the intersection of these three chip types and then normalize. >> Ben Bolstad has a page containing mixture cdfs for some arrays >> (http://bmbolstad.com/misc/mixtureCDF/MixtureCDF.html) but he never to my >> knowledge made one for the arrays you have. >> >> In general, I think you would be better off to normalize separately and then >> combine. > > Wouldn't another approach be to use one of them there meta-analysis > type approaches, such as: > > http://www.bioconductor.org/packages/2.10/bioc/html/RankProd.html or > http://www.bioconductor.org/packages/2.10/bioc/html/metaArray.html > > or? > > Never used them myself, but my mental note for these packages was > there for when I wanted to tackle such a situation (notes can be > wrong, of course -- mental ones, even more so since no ink was > actually invested). > > -steve > > -- > Steve Lianoglou > Graduate Student: Computational Systems Biology > ?| Memorial Sloan-Kettering Cancer Center > ?| Weill Medical College of Cornell University > Contact Info: http://cbio.mskcc.org/~lianos/contact -- Steve Lianoglou Graduate Student: Computational Systems Biology ?| Memorial Sloan-Kettering Cancer Center ?| Weill Medical College of Cornell University Contact Info: http://cbio.mskcc.org/~lianos/contact