[BioC] chip-chip data

Joern Toedling Joern.Toedling at curie.fr
Tue Mar 9 10:39:43 CET 2010


Hi,

they way I suggest to handle replicates is to combine them at the smoothing
step. The function "computeRunningMedians" has an argument "combineReplicates"
which you need to set to TRUE. And using the argument "modColumn" you specify
which column of your data object's phenoData marks the class and thus
indicates replicates.
Alternatively, you could take medians or means over the reporter-wise values
in each group but I think smoothing over replicates is preferable.

Regards,
Joern

On Mon, 8 Mar 2010 17:08:18 -0500, somnath bandyopadhyay wrote 
> Hello Joern, 
>   
> Thanks for all the suggestions. In my experiment, I had 3 replicates for WT
and 3 replicates for knockouts and each had respective inputs. Could you
please suggest me a way to use the replicate information? The examples you had
dinot have replicates in them. Any suggestions would be greatly appreciated. 
> Thanks, 
> Som. 
>   
>

--- 
 Joern Toedling 
 Institut Curie -- U900 
 26 rue d'Ulm, 75005 Paris, FRANCE 
 Tel. +33 (0)156246927



More information about the Bioconductor mailing list