Hi everybody, The human Agilent chip 4x44 has most of the probes only once, but some of then appear in 10 replicates. These replicates are genes and not controls. I am using the Limma package to analyze them. I would like to define these replicates as technical ones and leave the other probes (those that appear once) as they are. What is the best way to define only part of the spots as replicates? Thanks for your help Esti -- Ester Feldmesser, Ph.D. Bioinformatics Unit, Department of Biological Services Weizmann Institute of Science Levine Building, Room 110 phone: +972-8-934-2614 email: ester.feldmesser at weizmann.ac.il

Dear Ester, limma doesn't have a feature to treat just a few probes as duplicated. One reason for this omission is that unequal replication at the within-array level would play havoc with shrinking the variances. I'd just treat them all as independent probes. Best wishes Gordon >Date: Wed, 11 Jul 2007 16:14:54 +0300 >From: Ester Feldmesser <ester.feldmesser at="" weizmann.ac.il=""> >Subject: [BioC] Annalyzing replicates in the Agilent chip 4 x44 with > limma >To: bioconductor at stat.math.ethz.ch > >Hi everybody, > >The human Agilent chip 4x44 has most of the probes only once, but some >of then appear in 10 replicates. These replicates are genes and not >controls. >I am using the Limma package to analyze them. I would like to define >these replicates as technical ones and leave the other probes (those >that appear once) as they are. >What is the best way to define only part of the spots as replicates? > >Thanks for your help >Esti >-- > >Ester Feldmesser, Ph.D. >Bioinformatics Unit, Department of Biological Services >Weizmann Institute of Science >Levine Building, Room 110 >phone: +972-8-934-2614 >email: ester.feldmesser at weizmann.ac.il

Hi Ester, I do not think this is possible to do with Limma, short of doing a fair amount of hacking around in limma itself. In my experience the replicate probes in Agilent arrays behave almost identically (as opposed to different probes for a same gene). For simplicity's sake, we just use any one of them as representative. You could also take the average if you wanted. In my opinion, unless you see significant variations between those probes doing more than that would not be worth the effort. Francois On Wed, 2007-07-11 at 16:14 +0300, Ester Feldmesser wrote: > Hi everybody, > > The human Agilent chip 4x44 has most of the probes only once, but some > of then appear in 10 replicates. These replicates are genes and not > controls. > I am using the Limma package to analyze them. I would like to define > these replicates as technical ones and leave the other probes (those > that appear once) as they are. > What is the best way to define only part of the spots as replicates? > > Thanks for your help > Esti