Dear List, I have a tiem series experiment, I want to analyse with maSigPro. The data are from blood samples from 8 donors taken at 4 different timepoints under 2 treatments. My design matrix would look like this: > dm Time Replicate Control LPS I10R029a01.CEL 0 1 1 0 I10R029a03.CEL 0 1 1 0 I10R029a04.CEL 0 1 1 0 I10R029a05.CEL 0 1 1 0 I10R029b01.CEL 0 2 0 1 I10R029b03.CEL 0 2 0 1 I10R029b04.CEL 0 2 0 1 I10R029b05.CEL 0 2 0 1 I10R029c01.CEL 0 1 1 0 I10R029c03.CEL 0 1 1 0 I10R029c04.CEL 0 1 1 0 I10R029c05.CEL 0 1 1 0 I10R029d01.CEL 0 2 0 1 I10R029d05.CEL 0 2 0 1 I10R029d06.CEL 0 2 0 1 I10R029d07.CEL 0 2 0 1 I10R029e05.CEL 1 3 1 0 I10R029e06.CEL 1 3 1 0 I10R029e07.CEL 1 3 1 0 I10R029e08.CEL 1 3 1 0 I10R029f01.CEL 1 4 0 1 I10R029f05.CEL 1 4 0 1 I10R029f06.CEL 1 4 0 1 I10R029f07.CEL 1 4 0 1 I10R029g01.CEL 1 3 1 0 I10R029g02.CEL 1 3 1 0 I10R029g03.CEL 1 3 1 0 I10R029g04.CEL 1 3 1 0 I10R029h01.CEL 1 4 0 1 I10R029h02.CEL 1 4 0 1 I10R029h03.CEL 1 4 0 1 I10R029h04.CEL 1 4 0 1 I10R029i01.CEL 2 5 1 0 I10R029i02.CEL 2 5 1 0 I10R029i03.CEL 2 5 1 0 I10R029i04.CEL 2 5 1 0 I10R029j01.CEL 2 6 0 1 I10R029j02.CEL 2 6 0 1 I10R029j03.CEL 2 6 0 1 I10R029j04.CEL 2 6 0 1 I10R029k01.CEL 2 5 1 0 I10R029k02.CEL 2 5 1 0 I10R029k03.CEL 2 5 1 0 I10R029k04.CEL 2 5 1 0 I10R029l01.CEL 2 6 0 1 I10R029l02.CEL 2 6 0 1 I10R029l03.CEL 2 6 0 1 I10R029l04.CEL 2 6 0 1 I10R029m01.CEL 3 7 1 0 I10R029m02.CEL 3 7 1 0 I10R029m03.CEL 3 7 1 0 I10R029m04.CEL 3 7 1 0 I10R029n01.CEL 3 8 0 1 I10R029n02.CEL 3 8 0 1 I10R029n03.CEL 3 8 0 1 I10R029n04.CEL 3 8 0 1 I10R029o01.CEL 3 7 1 0 I10R029o02.CEL 3 7 1 0 I10R029o03.CEL 3 7 1 0 I10R029o04.CEL 3 7 1 0 I10R029p01.CEL 3 8 0 1 I10R029p02.CEL 3 8 0 1 I10R029p03.CEL 3 8 0 1 I10R029p04.CEL??? 3???????? 8?????? 0?? 1 However, since the samples were drawn from the same donors, I'd like to ask if there is a way to include this information also in the design matrix like in a paired analysis as in limma? Thanks for your input and kind regards, Michael

Dear Maria, Thanks for your answer. I allready thought that this might be your answer. In the meantime I run maSigPro with the design posted in the previous mail and the results look quite promising. However, yesterday I had the idea to simply subtract the t0 signal from the other time points for each donor. Recently, there was a posting from James Mac Donald I think, who showed that in limma this gives identical results as a paired analysis. Do you think that is approach might also be valid with maSigPro? Kind regards and a Merry Christmas, Mike -----Urspr?ngliche Nachricht----- Von: "M? Jos? Nueda" Gesendet: 23.12.2010 12:36:37 An: "Mike Walter" Betreff: Re: [BioC] Paired Data in maSigPro >Dear Mike, > >Thank you for using maSigPro and sorry for answering so late. >maSigPro does not take into account the subject (individual) information. We >would like to develop in the future this option to deal with longitudinal >data. >However I think that applying maSigPro with the design you show is correct >and it could be an adequate analysis. >Using longitudinal models requiere a lot of replicates to estimate the >model. > >Kind regards, > >Mar?a. > >-------------------------------------------------- >From: "Mike Walter" >Sent: Monday, December 13, 2010 1:43 PM >To: >Subject: [BioC] Paired Data in maSigPro > >> Dear List, >> >> I have a tiem series experiment, I want to analyse with maSigPro. The data >> are from blood samples from 8 donors taken at 4 different timepoints under >> 2 treatments. My design matrix would look like this: >> >>> dm >> Time Replicate Control LPS >> I10R029a01.CEL 0 1 1 0 >> I10R029a03.CEL 0 1 1 0 >> I10R029a04.CEL 0 1 1 0 >> I10R029a05.CEL 0 1 1 0 >> I10R029b01.CEL 0 2 0 1 >> I10R029b03.CEL 0 2 0 1 >> I10R029b04.CEL 0 2 0 1 >> I10R029b05.CEL 0 2 0 1 >> I10R029c01.CEL 0 1 1 0 >> I10R029c03.CEL 0 1 1 0 >> I10R029c04.CEL 0 1 1 0 >> I10R029c05.CEL 0 1 1 0 >> I10R029d01.CEL 0 2 0 1 >> I10R029d05.CEL 0 2 0 1 >> I10R029d06.CEL 0 2 0 1 >> I10R029d07.CEL 0 2 0 1 >> I10R029e05.CEL 1 3 1 0 >> I10R029e06.CEL 1 3 1 0 >> I10R029e07.CEL 1 3 1 0 >> I10R029e08.CEL 1 3 1 0 >> I10R029f01.CEL 1 4 0 1 >> I10R029f05.CEL 1 4 0 1 >> I10R029f06.CEL 1 4 0 1 >> I10R029f07.CEL 1 4 0 1 >> I10R029g01.CEL 1 3 1 0 >> I10R029g02.CEL 1 3 1 0 >> I10R029g03.CEL 1 3 1 0 >> I10R029g04.CEL 1 3 1 0 >> I10R029h01.CEL 1 4 0 1 >> I10R029h02.CEL 1 4 0 1 >> I10R029h03.CEL 1 4 0 1 >> I10R029h04.CEL 1 4 0 1 >> I10R029i01.CEL 2 5 1 0 >> I10R029i02.CEL 2 5 1 0 >> I10R029i03.CEL 2 5 1 0 >> I10R029i04.CEL 2 5 1 0 >> I10R029j01.CEL 2 6 0 1 >> I10R029j02.CEL 2 6 0 1 >> I10R029j03.CEL 2 6 0 1 >> I10R029j04.CEL 2 6 0 1 >> I10R029k01.CEL 2 5 1 0 >> I10R029k02.CEL 2 5 1 0 >> I10R029k03.CEL 2 5 1 0 >> I10R029k04.CEL 2 5 1 0 >> I10R029l01.CEL 2 6 0 1 >> I10R029l02.CEL 2 6 0 1 >> I10R029l03.CEL 2 6 0 1 >> I10R029l04.CEL 2 6 0 1 >> I10R029m01.CEL 3 7 1 0 >> I10R029m02.CEL 3 7 1 0 >> I10R029m03.CEL 3 7 1 0 >> I10R029m04.CEL 3 7 1 0 >> I10R029n01.CEL 3 8 0 1 >> I10R029n02.CEL 3 8 0 1 >> I10R029n03.CEL 3 8 0 1 >> I10R029n04.CEL 3 8 0 1 >> I10R029o01.CEL 3 7 1 0 >> I10R029o02.CEL 3 7 1 0 >> I10R029o03.CEL 3 7 1 0 >> I10R029o04.CEL 3 7 1 0 >> I10R029p01.CEL 3 8 0 1 >> I10R029p02.CEL 3 8 0 1 >> I10R029p03.CEL 3 8 0 1 >> I10R029p04.CEL 3 8 0 1 >> >> However, since the samples were drawn from the same donors, I'd like to >> ask if there is a way to include this information also in the design >> matrix like in a paired analysis as in limma? >> >> Thanks for your input and kind regards, >> >> Michael >> >> _______________________________________________ >> Bioconductor mailing list >> Bioconductor at r-project.org >> https://stat.ethz.ch/mailman/listinfo/bioconductor >> Search the archives: >> http://news.gmane.org/gmane.science.biology.informatics.conductor >