Hi,
I have a quick question for you, not sure if you are able to help if not can you point me towards the correct person.
I am following this tutorial: Group-specific condition effects, individuals nested within groups and I understand that I have to create some sort of dummy variable to account for the nestedness of my design. So I was able to do that but I am unsure if I did it correctly since I still get the error stating: model not full rank.
My question is, I have 9 plots (6 burned and 3 unburned) and each plot has 3 subplots where I am sampling over time. So, therefore, the subplots are nested within the plots so following the tutorial I added a dummy based on the subplot, but since this is also a time series, do I need to add a dummy for Plot since I have multiples of it given the time series? Would this be correct?
so something like this, where pl.in and sub.in are the dummy variables?
plot | pl.in | subplot | sub.in | Treatment
1 | 1 | 1 | 1 | A
1 | 1 | 2 | 2 | A
2 | 2 | 1 | 1 | A
2 | 2 | 2 | 2 | A
3 | 3 | 1 | 1 | A
1 | 1 | 1 | 1 | B
1 | 1 | 2 | 2 | B
2 | 2 | 1 | 1 | B
2 | 2 | 2 | 2 | B
3 | 3 | 1 | 1 | B
I know you are quite busy, but I cannot seem to locate this information online, on this forum and or on edger (when I google), if you could please help I would really appreciate it as I really want to use Deseq for this analysis.
You have mentioned three different packages: edgeR, DESeq and DESeq2. Do you understand that these are all different packages developed by different people?
edgeR has a substantial User's Guide and many workflows online. Almost all the experimental design advice in the limma User's Guide also applies to edgeR. It's not clear that you have accessed any of these documentation sources yet.