Analysing a timecourse/loop experiment without replication with limma
1
0
Entering edit mode
Claus Mayer ▴ 340
@claus-mayer-1179
Last seen 7.1 years ago
European Union
Dear all! I am trying to analyse a timecourse experiment using two-color microarrays with limma. There are 5 timepoints compared in a loop-design, i.e 1 vs 2, 2 vs 3,?, 4 vs 5, 5 vs 1. This experiment is supposed to be repeated 3 more times (with independent samples) but my collaborators were interested to see the results for this first set of experiments before they hybridise the rest. I have put the data into limmaGUI and chose the 4 first present comparisons (2 minus 1, 3 minus 2, 4 minus 3 and 5 minus 4) as parameters. When I try to fit the model I get an error message saying ?Error in colSums(fit$effects(fit$rank+1):narrays,]^2 ?x? must be an array of at least two dimensions?. At first I thought I just ran into a singularity problem caused by the lack of repitition (although I had originally thought with each sample being present twice that should be no problem). But then I tried the same thing with a completely different data set, that had the identical design (with 4 instead of 5 timepoints). In this case everything went fine, which indicates something else must have been wrong with my first analysis. So I went pack and just changed the targets file for my original data set and now claimed that each array was just comparing a treatment with control. The reasoning was to check whether there was anything wrong the data files themselves. With this new (and completely wrong) design I was able to run the analysis, which means that there is nothing obviously wrong with the data. I probably have overlooked something very obvious or I am just doing something stupid but perhaps one of you can put me in the right direction. Any helpful comments appreciated. Thanks, Claus -- ********************************************************************** ************* Claus-D. Mayer | http://www.bioss.ac.uk Biomathematics & Statistics Scotland | email: claus at bioss.ac.uk Rowett Research Institute | Telephone: +44 (0) 1224 716652 Aberdeen AB21 9SB, Scotland, UK. | Fax: +44 (0) 1224 715349
0
Entering edit mode
Naomi Altman ★ 6.0k
@naomi-altman-380
Last seen 6 months ago
United States
A single loop design includes 2 replicates. I would do a separate channel analysis as described in the Limma manual section 17. --Naomi At 02:12 PM 8/19/2005, Claus Mayer wrote: >Dear all! > >I am trying to analyse a timecourse experiment using two-color >microarrays with limma. There are 5 timepoints compared in a >loop-design, i.e 1 vs 2, 2 vs 3, , 4 vs 5, 5 vs 1. This experiment is >supposed to be repeated 3 more times (with independent samples) but my >collaborators were interested >to see the results for this first set of experiments before they >hybridise the rest. I have put the data into limmaGUI and chose the 4 >first present comparisons (2 minus 1, 3 minus 2, 4 minus 3 and 5 minus >4) as parameters. When I try to fit the model I get an error message >saying ?Error in colSums(fit$effects(fit$rank+1):narrays,]^2 ?x? must be >an array of at least two dimensions?. > >At first I thought I just ran into a singularity problem caused by the >lack of repitition (although I had originally thought with each sample >being present twice that should be no problem). But then I tried the >same thing with a completely different data set, that had the identical >design (with 4 instead of 5 timepoints). In this case everything went >fine, which indicates something else must have been wrong with my first >analysis. > >So I went pack and just changed the targets file for my original data >set and now claimed that each array was just comparing a treatment with >control. The reasoning was to check whether there was anything wrong the >data files themselves. With this new (and completely wrong) design I was >able to run the analysis, which means that there is nothing obviously >wrong with the data. > >I probably have overlooked something very obvious or I am just doing >something stupid but perhaps one of you can put me in the right direction. > >Any helpful comments appreciated. > >Thanks, > > >Claus > >-- >********************************************************************* ************** > Claus-D. Mayer | http://www.bioss.ac.uk > Biomathematics & Statistics Scotland | email: claus at bioss.ac.uk > Rowett Research Institute | Telephone: +44 (0) 1224 716652 > Aberdeen AB21 9SB, Scotland, UK. | Fax: +44 (0) 1224 715349 > >_______________________________________________ >Bioconductor mailing list >Bioconductor at stat.math.ethz.ch >https://stat.ethz.ch/mailman/listinfo/bioconductor Naomi S. Altman 814-865-3791 (voice) Associate Professor Bioinformatics Consulting Center Dept. of Statistics 814-863-7114 (fax) Penn State University 814-865-1348 (Statistics) University Park, PA 16802-2111