[BioC] CGHcall and Normalization

Salih Tuna st5 at sanger.ac.uk
Mon Jun 13 18:19:00 CEST 2011


Thanks a lot Martin. That is very helpful.
Best,
salih

-----Original Message-----
From: Martin Morgan [mailto:mtmorgan at fhcrc.org] 
Sent: 13 June 2011 17:04
To: Salih Tuna
Cc: bioconductor at r-project.org
Subject: Re: [BioC] CGHcall and Normalization

On 06/13/2011 08:01 AM, Salih Tuna wrote:
> Hi Martin,
> Thanks for your reply. I managed to run the code. AS you said I needed
> to adjust the data.
> But now the output seems something like this:
>
> normalized.data
> cghRaw (storageMode: lockedEnvironment)
> assayData: 3552 features, 5 samples
>    element names: copynumber
> phenoData
>    sampleNames: AdCA10, SCC27, ..., SCC39  (5 total)
>    varLabels and varMetadata description: none
> featureData
>    featureNames: RP11-465B22, RP4-785P20, ..., CTB-99K24  (3552 total)
>    fvarLabels and fvarMetadata description:
>      Chromosome: Chromosomal position
>      Start: Basepair position start
>      End: Basepair position end
> experimentData: use 'experimentData(object)'
> Annotation:
>
> This might be a very stupid question but I was expecting to see the
> normalized data which is the same dimension as the raw data. I am sure
I
> can access the normalized data which is the same dimension as the raw
> data. But I don't know how. Could you please show me how to do this?

The docs for ?normalize say that the 'value' returned is a data.frame, 
but...

class(normalized.data) tells me this is a cghRaw object (as does the 
display, above). So I looked at ?cghRaw and found that I could extract 
the copy number data with

   copynumber(normalized.data)

And it seems to be different from (i.e., normalized) the raw data

 > example(CGHcall)
[snip]
 > colSums(copynumber(raw.data))
     AdCA10      SCC27      SCC32      SCC36      SCC39
-215.51938   46.58693 -153.71559  398.60994  518.95828
 > colSums(copynumber(normalized.data))
     AdCA10      SCC27      SCC32      SCC36      SCC39
-127.75115  -39.93650  -69.26762  134.13213  409.15257

so the documentation (also browseVignettes(package="CGHcall")) seems to 
be your friend (if a little difficult to navigate).

Martin

>
> Best,
> salih
>
> -----Original Message-----
> From: Martin Morgan [mailto:mtmorgan at fhcrc.org]
> Sent: 11 June 2011 16:24
> To: Salih Tuna
> Cc: bioconductor at r-project.org
> Subject: Re: [BioC] CGHcall and Normalization
>
> On 06/08/2011 08:47 AM, Salih Tuna wrote:
>> Hi,
>>
>> i am trying to use the normalization function from CGHcall library.
>>
>> The command i use is
>>
>>
>>
>> normal.fullData<- normalize(fullData, method = "median", cellularity
=
>> 1)
>>
>> and i get the following error message. How can i solve this issue.
>>
>> Error in function (classes, fdef, mtable)  :
>>
>>     unable to find an inherited method for function "copynumber", for
>> signature "data.frame"
>
> Hi Salih --
>
> If I
>
>     example(CGHcall)
>
> and look at how normalize() is used, the first argument is 'raw.data'
> and it is
>
>   >  class(raw.data)
> [1] "cghRaw"
> attr(,"package")
> [1] "CGHbase"
>
> if instead I
>
>   >  normalize(data.frame())
> Applying median normalization ...
> Error in function (classes, fdef, mtable)  :
>     unable to find an inherited method for function "copynumber", for
> signature "data.frame"
>
> so perhaps you are trying to use a data.frame instead of an object of
> 'cghRaw', i.e., your earlier work flow needs to be adjusted. If you
need
>
> more help, it is best to post a reproducible example, starting from
> sample data available in the package.
>
> Hope that helps,
>
> Martin
>>
>> best,
>>
>> salih
>>
>>
>>
>>
>
>


-- 
Computational Biology
Fred Hutchinson Cancer Research Center
1100 Fairview Ave. N. PO Box 19024 Seattle, WA 98109

Location: M1-B861
Telephone: 206 667-2793


-- 
 The Wellcome Trust Sanger Institute is operated by Genome Research 
 Limited, a charity registered in England with number 1021457 and a 
 company registered in England with number 2742969, whose registered 
 office is 215 Euston Road, London, NW1 2BE.



More information about the Bioconductor mailing list