[BioC] ANOVA in edgeR?

[Gordon K Smyth]

Dear Vincenzo,

edgeR does use what can be viewed as a generalization of ANOVA, but one 
that is appropriate for count data.  It does not use the ANOVA formula 
that you may have seen in an introductory statistics textbook, which are 
only for normally distributed data, and only for simple types of 
experiments.

MDS in edgeR is not a sort of SSB.

Best wishes
Gordon

> Date: Mon, 1 Oct 2012 11:25:20 +0200
> From: Vincenzo Capece <vivo0304 at gmail.com>
> To: Natasha Sahgal <nsahgal at well.ox.ac.uk>
> Cc: "bioconductor at r-project.org" <bioconductor at r-project.org>
> Subject: Re: [BioC] DESeq - working with more than 2 conditions
>
> Thanks a lot Natasha, edgeR is exactly what I was looking for.
> I am also a statistics beginner and I have an another question ,but I don't
> know if it's correct or not: To compute BCV in edgeR, why the developers
> didn't use ANOVA? I think that the MDS in edgeR is s sort  of SSB(sum of
> squares between) in ANOVA.
> Thank you very much.
>
> On 13 September 2012 10:46, Natasha Sahgal <nsahgal at well.ox.ac.uk> wrote:
>
>> Dear Vincenzo,
>>
>> You could try to use the edgeR package (and the glm method within it).
>> Also, if I am not mistaken, you could use DESeq but you have to use the
>> glm method there.
>>
>>
>> Hope that helps,
>> Natasha
>>
>>
>> -----Original Message-----
>> From: bioconductor-bounces at r-project.org [mailto:
>> bioconductor-bounces at r-project.org] On Behalf Of Vincenzo Capece
>> Sent: 13 September 2012 07:53
>> To: bioconductor at r-project.org
>> Subject: [BioC] DESeq - working with more than 2 conditions
>>
>> Dear all,
>> I am a R beginner and I hope you can help me.
>> I have a question about DESeq: DESeq works fine if I have 2
>> conditions(treated,untreated)
>> Example:
>>                                           treated   untreated
>> ENSMUSG00000000001      193     254
>> ENSMUSG00000000003      0       0
>> ENSMUSG00000000028      10      13
>> ENSMUSG00000000031      0       4
>> ENSMUSG00000000037      8       9
>> ENSMUSG00000000049      6       8
>> ...
>> ...
>> ...
>>
>>
>> but if I have data about more than two conditions, for instance
>> 1-month-treated,3-months-treatead, 6-months-treated,untreated, can I use
>> DESeq? If not, Which R package (if it exists)do you suggest to use?
>>
>> Thanks in advance.
>> Regards.
>
> -- 
> Regards,
>            Capece Vincenzo

______________________________________________________________________
The information in this email is confidential and intend...{{dropped:4}}