[BioC] bug in limma?

[Simon Melov]

OK James,
Version 1.3.14. I subset'd MA$genes by itself, tried fit, still same  
error message. I then subset'd MA$M and MA$A just in case, no change -  
still same error message. I then regenerated the GAL file as a CSV file  
to be sure there were no idiosyncrasies with the way our gal files are  
generated, same error message on fit.

One other thing, I am reading in more files than I am using in my  
design (i.e. I'm reading in a 100 arrays, but am only using 52 of them  
in the design - the remainder are for another design). Could this be  
the source of the error? Does limma require all arrays in the initial  
read-in to be used in the design?

thanks

S.

On Jan 29, 2004, at 7:37 PM, James Wettenhall wrote:

> Simon,
>
> It would be helpful to know which version of limma you are
> using and how you read in your genelist.  Did you use
> RG$genes <- readGAL()  ?
>
> Do you have any blank lines at the end of your GAL file?
>
> Have a look at the help for readGAL and read.table (which
> readGAL calls).
>
> What happens if you subset your genes component of MA (or RG),
> then does lmFit work?
>
> i.e.
>
> MA$genes <- MA$genes[1:16896,]
>
> Regards,
> James
>
> On Thu, 29 Jan 2004, Simon Melov wrote:
>
>> Hi,
>> this may be my ignorance, but I am persistently getting an error upon
>> reading in one data set. It seems to be in the implementation of the
>> gene list in the normalized MA.
>>
>> The error I am getting after trying to fit is..
>>
>>> fit <- lmFit(MA, design)
>> Error: (subscript) logical subscript too long
>>
>> After checking the size of M,A, and the gene list in MA, its clear  
>> that
>> M and A are the correct size. But for some reason, the genelist is
>> incorrectly sized however you look at it (which is not the case in the
>> input gpr files..)
>>
>> show(MA)
>> An object of class "MAList"
>> $genes
>>    Block Row Column                    ID                  Name
>> 1     1   1      1           Human Cot 1           Human Cot 1
>> 2     1   1      2 Spot Report Product 2 Spot Report Product 2
>> 3     1   1      3        B-Human Actin         B-Human Actin
>> 4     1   1      4                Buffer                Buffer
>> 5     1   1      5                Poly A                Poly A
>> 16891 more rows ...
>>
>>> MA$M[16891:16896,]
>>           MB12A04     MB12A05  ect..
>> [1,] -0.08841763 -0.24117406
>> [2,] -0.01282521 -0.09876827
>> [3,]  0.16414844  0.44696644
>> [4,]  0.26038965  0.51199616
>> [5,]  0.28553731 -0.51235751
>> [6,]  0.03343551  0.31066801
>>
>>> MA$M[16894:16897,]
>> Error: subscript out of bounds
>>
>> This is fine, as there are only 16896 rows.
>>
>> But for the MA$genes..
>>
>> MA$gen[16894:16897,]
>>        Block Row Column                    ID                  Name
>> 16894    32  22     22 Spot Report Product 8 Spot Report Product 8
>> 16895    32  22     23                Buffer                Buffer
>> 16896    32  22     24                Buffer                Buffer
>> NA       NA  NA     NA                  <NA>                  <NA>
>>
>>   MA$gen[16894:16898,]
>>        Block Row Column                    ID                  Name
>> 16894    32  22     22 Spot Report Product 8 Spot Report Product 8
>> 16895    32  22     23                Buffer                Buffer
>> 16896    32  22     24                Buffer                Buffer
>> NA       NA  NA     NA                  <NA>                  <NA>
>> NA.1     NA  NA     NA                  <NA>                  <NA>
>>
>> MA$gen[16894:16899,]
>>        Block Row Column                    ID                  Name
>> 16894    32  22     22 Spot Report Product 8 Spot Report Product 8
>> 16895    32  22     23                Buffer                Buffer
>> 16896    32  22     24                Buffer                Buffer
>> NA       NA  NA     NA                  <NA>                  <NA>
>> NA.1     NA  NA     NA                  <NA>                  <NA>
>> NA.2     NA  NA     NA                  <NA>                  <NA>
>>
>> and so on.
>>
>> Is this why I'm getting the fit error listed above, ie is this a bug?
>>
>> Simon
>
> --  
> ----------------------------------------------------------------------- 
> ---
> James Wettenhall                                  Tel: (+61 3) 9345  
> 2629
> Division of Genetics and Bioinformatics           Fax: (+61 3) 9347  
> 0852
> The Walter & Eliza Hall Institute         E-mail:  
> wettenhall at wehi.edu.au
>  of Medical Research,                     Mobile: (+61 / 0 ) 438 527  
> 921
> 1G Royal Parade,
> Parkville, Vic 3050, Australia
> http://www.wehi.edu.au
> ----------------------------------------------------------------------- 
> ---
>
>
Simon Melov Ph.D
Associate Professor &
Director of Genomics
Buck Institute for Age Research
8001 Redwood Blvd
Novato, CA 94945
415 209 2068
Fax: 415 209 2231