[BioC] reading imagene files with LIMMA

[Martin Morgan]

Yi Xing --

What does sessionInfo() say? I suspect you are using an older version
of limma, and that the bug you are encountering was fixed more than a
year ago.

Download and install the latest release of R from

http://cran.r-project.org

and use

> source('http://bioconductor.org/biocLite.R')
> biocLite()

to get back on track.

Best,

Martin

Yi Xing <yxing at ucla.edu> writes:

> Hello Saroj,
>
> Thanks for your message. I think the problem is with the Field 
> Dimensions. It looks like this for my file:
>
> Field     Metarows Metacols Rows Cols
>    Block1         1        1   15   14
>    Block2         1        1   15   14
>    Block3         1        1   15   14
>    Block4         1        1   15   14
>    Block5         1        1   15   14
>    Block6         1        1   15   14
>    Block7         1        1   15   14
>    Block8         1        1   15   14
>    Block9         1        1   15   14
>    Block10        1        1   15   14
>    Block11        1        1   15   14
>    Block12        1        1   15   14
>    Block13        1        1   15   14
>    Block14        1        1   15   14
>    Block15        1        1   15   14
>    Block16        1        1   15   14
>    Block17        1        1   15   14
>    Block18        1        1   15   14
>    Block19        1        1   15   14
>    Block20        1        1   15   14
>    Block21        1        1   15   14
>    Block22        1        1   15   14
>    Block23        1        1   15   14
>    Block24        1        1   15   14
>    Block25        1        1   15   14
>    Block26        1        1   15   14
>    Block27        1        1   15   14
>    Block28        1        1   15   14
>    Block29        1        1   15   14
>    Block30        1        1   15   14
>    Block31        1        1   15   14
>    Block32        1        1   15   14
>
> so read.imagene() calculates the total number of spots as prod(FD):
>
> nspots <- prod(FD)
> Y <- matrix(0, nspots, narrays)
>
>   which is 2.046527e+74, causing the error in the next line.
>
> Yi
> On Jun 28, 2007, at 2:21 PM, Saroj Mohapatra wrote:
>
>> Hello Yi,
>>
>> It is possible that there is a discrepancy between the number of spots 
>> calculated from the Field Dimensions in header and the number of rows 
>> it actually reads from the file.
>>
>> For example, in my Imagene 5.1 file:
>> in the header, below "Begin Field Dimensions" I find the following:
>> Metarows = 8
>> Metacols = 4
>> Rows = 15
>> Cols = 12
>>
>> So, the total number of spots = 5760
>> In the data section of my file, there are 5760 rows (excluding column 
>> titles).
>>
>> Of course, the actual number would vary, you would have a different 
>> number of spots. You could start with checking that the numbers match.
>>
>> Also, there should be something like "Begin Raw Data" before the first 
>> line and "End Raw Data" after the last line of data section.
>>
>> Good luck!
>>
>> Saroj
>>
>> Yi Xing wrote:
>>
>>> Hi,
>>>
>>> I tried to read several Imagene files with LIMMA:
>>>
>>> library(limma)
>>> targets<-readTargets()
>>> files<-targets[,c("FileNameCy3","FileNameCy5")]
>>> RG<-read.maimages(files,source='imagene')
>>>
>>> however, I got this error message:
>>>
>>> Read header information
>>> Error in matrix(0, nspots, narrays) : matrix: invalid 'nrow' value 
>>> (too large or NA)
>>> In addition: Warning message:
>>> NAs introduced by coercion
>>>
>>> What might be the problem and how to fix it? Thanks.
>>>
>>> Yi Xing
>>>
>>> _______________________________________________
>>> Bioconductor mailing list
>>> Bioconductor at stat.math.ethz.ch
>>> https://stat.ethz.ch/mailman/listinfo/bioconductor
>>> Search the archives: 
>>> http://news.gmane.org/gmane.science.biology.informatics.conductor
>>>
>> <smohapat.vcf>
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor

-- 
Martin Morgan
Bioconductor / Computational Biology
http://bioconductor.org