Difference Between Oligo And Affy Packages For Affymetrix Gene St Array Pre-Processing

Question

[Adam Cornwell]

Hello, I've got some Affymetrix Gene ST arrays I'm working with. My usual workflow is with Affy/Limma for Affymetrix 3' arrays, but I always see the Gene ST arrays brought up with Oligo and Aroma for pre-processing. Does the choice between Affy and Oligo actually make any difference in results, if I'm only using it for reading in arrays, basic QA graphs, and RMA? I've used both Affy and Oligo on a dataset previously a couple of years ago, and didn't really see any difference in the results that I can remember, so I'm still trying to figure out why everyone seems to use Oligo and Aroma for the Gene ST arrays.

Bonus question: I've been using FARMS quite a bit on 3' arrays for the I/NI filtering functionality, as well as their quantile normalization method. Overall I know the underlying structure of the design for the 3' arrays and the Gene ST arrays isn't all that different, but I'm afraid there might be some different different assumptions that I'm not aware of that could preclude the application of some normalization methods. Is it reasonable to use FARMS with Gene ST arrays?

Any insights appreciated!

[Adam Cornwell]

I suppose I'll take the silence to mean "look at the code" or "post to the BioC list".

Answer 1

[steve.s.sheridan]

In general, use oligo unless it doesn't work, in which case, try affy. If either of them works, you're fine; the differences between them are pretty small.

I've just posted more information on this topic here.