Hi, 

I'm basically trying to follow up a post I found - https://groups.google.com/forum/#!topic/combat-user-forum/PcTxNlaUmAI

Ultimately, I'm trying to perform a batch correction on Microarray data with two known, and different technical effects (batches). 

I realise this can be accounted for using an additive model, but I'm interested on how the ComBat function in SVA could be used to tackle such a problem.

I'm wondering which option is the "done thing":

Apply combat twice - Apply once on Technical effect A, then apply resultant expression set on Technical effect B

Merge All Technical Effects - Regardless of origin, merge Technical Effect A and B into one. i.e, Technical Effect A (two batches; 1 and 2), Technical Effect B (two batches; 1 and 2) - Inform ComBat of 4 Batches.

Any suggestions??

Thanks!

Hi Andrew,

I guess combining the two batches into one factor with levels 1:1, 1:2, 2:2, and 2:2 is the best option if:

(a) you have enough samples in each of the four batch group

(b) you have all / most condtions in each batch group.

A two steps approach might not work so well, if either (a) or (b) is not the case, you can try SVA or similar, or an

algorithm that explicitly models correlations, like:

http://cran.r-project.org/web/packages/Tsphere/index.html

or the cat score in

http://cran.r-project.org/web/packages/st/index.html

In my experience, theses work very well if  batch effects are present.

In any case, if you decide to correct the data, check the clustering with heatmaps, PCA etc.

Best wishes,

Bernd