Dear Gordon & list Why is it not possible to fit random block effect and technical replicates within array in the same linear model? Is it because - the statistical theory is not ready? - or - it is not implemented in Limma? I am not a statistician so it might be a stupid question :-) Thanks Jakob ------------------------------------------------------------------- Jakob Hedegaard Danish Institute of Agricultural Sciences Department of Genetics and Biotechnology Molecular Genetics and System Biology Building K25 Research Centre Foulum P.O. box 50 DK-8830 Tjele Denmark Tel: (+45)89991363 Fax: (+45)89991300 -----Oprindelig meddelelse----- Fra: bioconductor-bounces@stat.math.ethz.ch [mailto:bioconductor- bounces@stat.math.ethz.ch] P? vegne af Gordon K Smyth Sendt: Thursday, June 02, 2005 12:43 PM Til: Peter S?rensen (HAG) Cc: bioconductor@stat.math.ethz.ch Emne: [BioC] random block effect and technical replicates within array at the same in linear model using Limma? > Date: Thu, 2 Jun 2005 11:47:15 +0200 > From: Peter S?rensen (HAG) <peter.sorensen2@agrsci.dk> > Subject: [BioC] random block effect and technical replicates within > array at the same in linear model using Limma ? > To: <bioconductor@stat.math.ethz.ch> > > Dear list, > > I have an experiment with 24 arrays. > I have 3 factors each with 2 levels and 12 blocks (1,1,2,2,3,3,....,12,12). > Each gene is replicated 4 times on each array. > > I know how to fit blocks as random as explained in the Limma manual (Technical replication I) > I know how to account for the replication within arrays. > > > Is it possible to fit random block effect and technical replicates within array at the same in > linear model using Limma ? No it isn't. (Being able to do so awaits the development of new theory.) See ?duplicateCorrelation 2nd paragraph of "Details:", or ?lmFit 4th paragraph of "Details:". Gordon > Thanks > Peter _______________________________________________ Bioconductor mailing list Bioconductor@stat.math.ethz.ch https://stat.ethz.ch/mailman/listinfo/bioconductor

On Wed, June 8, 2005 4:11 am, Jakob Hedegaard said: > Dear Gordon & list > > Why is it not possible to fit random block effect and technical replicates within array in the > same linear model? > Is it because > - the statistical theory is not ready? - or > - it is not implemented in Limma? The first reason. This is why I said "Being able to do so awaits the development of new theory". Theory already exists of course to fit linear models with more than one random effect. The aspect of the theory which is not ready is an effective way to do the between-gene smoothing. Gordon > I am not a statistician so it might be a stupid question :-) > > Thanks > Jakob > > ------------------------------------------------------------------- > Jakob Hedegaard > Danish Institute of Agricultural Sciences > Department of Genetics and Biotechnology > Molecular Genetics and System Biology > Building K25 > Research Centre Foulum > P.O. box 50 > DK-8830 Tjele > Denmark > Tel: (+45)89991363 > Fax: (+45)89991300 > -----Oprindelig meddelelse----- > Fra: bioconductor-bounces@stat.math.ethz.ch [mailto:bioconductor- bounces@stat.math.ethz.ch] P? > vegne af Gordon K Smyth > Sendt: Thursday, June 02, 2005 12:43 PM > Til: Peter S?rensen (HAG) > Cc: bioconductor@stat.math.ethz.ch > Emne: [BioC] random block effect and technical replicates within array at the same in linear model > using Limma? > >> Date: Thu, 2 Jun 2005 11:47:15 +0200 >> From: Peter S?rensen (HAG) <peter.sorensen2@agrsci.dk> >> Subject: [BioC] random block effect and technical replicates within >> array at the same in linear model using Limma ? >> To: <bioconductor@stat.math.ethz.ch> >> >> Dear list, >> >> I have an experiment with 24 arrays. >> I have 3 factors each with 2 levels and 12 blocks (1,1,2,2,3,3,....,12,12). >> Each gene is replicated 4 times on each array. >> >> I know how to fit blocks as random as explained in the Limma manual (Technical replication I) >> I know how to account for the replication within arrays. >> >> >> Is it possible to fit random block effect and technical replicates within array at the same in >> linear model using Limma ? > > No it isn't. (Being able to do so awaits the development of new theory.) See > ?duplicateCorrelation 2nd paragraph of "Details:", or ?lmFit 4th paragraph of "Details:". > > Gordon > >> Thanks >> Peter