[BioC] PCA to remove batch effect.

[Fabrice Tourre]

Kasper,

I know if the batch effect is known. We can use Combat to get the new
expression index without batch effect. But if the batch effect is
unknown, it seems that sva can not give the new expression index
without batch effect. In the example of SVA document, sva first get
svobj$sv (surrogate variables), then using it in the f.test. How can
we output the new expression index without batch effect? Can we use
PCs(principal components) to replace surrogate variables?

Sorry for this basic questions.


On Wed, Dec 14, 2011 at 12:49 AM, Kasper Daniel Hansen
<kasperdanielhansen at gmail.com> wrote:
> ComBat is now part of the sva package, so it hosts both batch
> correction effects.
>
> Kasper
>
> On Tue, Dec 13, 2011 at 11:35 AM, Richard Friedman
> <friedman at cancercenter.columbia.edu> wrote:
>> Dear Fabrice,
>>
>>        As I recall you label the input samples by batches (dates of
>> processing)
>> and the program takes care of the rest.
>>
>> Best wishes,
>> Rich
>>
>>
>> On Dec 13, 2011, at 11:33 AM, Fabrice Tourre wrote:
>>
>>> When using ComBat, it need to know the batch effect first.
>>>
>>> Thanks.
>>>
>>> On Wed, Dec 14, 2011 at 12:29 AM, Richard Friedman
>>> <friedman at cancercenter.columbia.edu> wrote:
>>>>
>>>> Rich
>>
>>
>> _______________________________________________
>> Bioconductor mailing list
>> Bioconductor at r-project.org
>> https://stat.ethz.ch/mailman/listinfo/bioconductor
>> Search the archives:
>> http://news.gmane.org/gmane.science.biology.informatics.conductor