Dear BC community,
As you might already know immgen is a public database that hosts microarray data from various immune cell populations. There are more than 200 cell types in there (with raw data coming from 2-3 biological replicates that are accessible in this and this links).
I'd like crunch all these data together to come up with a list of significantly and differentially expressed genes that discriminate the cell types from one another. I'm planning to use this information in a separate pipeline I'm working on. I started following the vignettes of packages like limma, but I wanted to get your opinions whether there is a preferred method of approaching a multidimensional dataset like this. It is a bit different than straightforward 2 or 3 group comparisons. Would limma work in my case? An important factor to consider here may be the numbers of biological replicates per cell type (some samples have 2 and most others have 3). If you have suggestions for approaching this kind of data, please let me know.
Best,
Atakan
