[BioC] Merging different datasets?

Sean Davis sdavis2 at mail.nih.gov
Thu Oct 28 20:16:02 CEST 2004


If you know that your slides have the same clones/probes and the 
reference, etc. is the same, you could directly combine the RG objects. 
  You just need to have the "rows" from the RG objects matching up.  So, 
if you have a unique id in the genes slot of the RG object, you could 
use the smaller group to get only those genes from the larger doing 
something like:

RG.larger2smaller <- RG.larger[as.character(RG.larger$genes$uniqueid) 
%in% as.character(RG.smaller$genes$uniqueid),]

The idea here is that for all genes where the unique ID in the larger 
set is in the smaller set, keep that gene.  Otherwise, that gene will 
not be kept.  RG.larger2smaller will have then have the same number of 
genes as the RG.smaller.  Then, you will have to make sure they are 
ordered in the same way.

Sean

On Oct 28, 2004, at 12:22 PM, Silvano Piazza wrote:

> Hi everyones,
>
> I have two different datasets of cDNA microarray experiments regarding 
> the same kind of samples, but two different kind of slices, in the 
> sense that  dataset1 was hybridized with slides with set of genes, and 
> dataset2 was hybridized with slides with set of the sames genes(so ... 
> clone phisical identity) PLUS other genes.
>
> So now I would like to merge the results from the two(of course 
> only,for the common genes). My idea was
> 1) load each dataset alone and  normalize it (I am using limma)
>    That's OK
> 2) merge the 2 datasets with MergeMaid.
>    That's NOT OK because first of all MergeMaid need ExprSet class 
> object in input, and using as(RG,"ExprSet") function from Biobase
> create ExprSet2 class object.
> Second because in this way I can import only RG class object and not 
> MA class object, but I suppose I HAVE to normailzed each dataset on 
> its own!
>
> So could anyone help me?
> Maybe, I miss the right way to use MargeMaid...
>
> Or anyway there is another methods the better marge datasets with 
> common genes?
>
> Thank you,everyone!
>
> Silvano
>
>
> Silvano Piazza
> LNCIB Trieste,
> Area Science Park
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor



More information about the Bioconductor mailing list