[BioC] rbind of MALists

[Gordon K Smyth]

> Date: Mon, 12 Jun 2006 15:20:24 -0700
> From: "Matthew Scholz" <schmatthew at gmail.com>
> Subject: [BioC] rbind of MALists
> To: bioconductor at stat.math.ethz.ch
>
> Hi all,
>
> I'm trying to combine two MALists, named RGAmodel and RGBmodel, using
> rbind and am running into the following problem:
>
>> rbind(RGAmodel, RGBmodel)
> Error in match.names(clabs, names(xi)) : names don't match previous names:
>         Name
> In addition: Warning message:
> longer object length
>         is not a multiple of shorter object length in: clabs == nmi
>
> I checked that the names were the same for each object and that the
> classes and dimensions match:
>
>> names(RGAmodel)
> [1] "weights"     "targets"     "genes"       "source"      "printer"
> [6] "genesStatus" "M"           "A"
>> names(RGBmodel)
> [1] "weights"     "targets"     "genes"       "source"      "printer"
> [6] "genesStatus" "M"           "A"
>
>> class(RGAmodel)
> [1] "MAList"
> attr(,"package")
> [1] "limma"
>
>> class(RGBmodel)
> [1] "MAList"
> attr(,"package")
> [1] "limma"
>
>> dim(RGAmodel)
> [1] 32448    32
>
>> dim(RGBmodel)
> [1] 32448    32
>
> What am I doing wrong? I'm using R2.3.0 and Bioconductor 1.7
>
> Thanks,
>
> Matt
>
> --
> Matt Scholz
> Senior Research Specialist
> Department of Plant Sciences
> University of Arizona
> (520) 621-1695

Dear Matt,

Firstly, your MAList objects contain non-standard components "genesStatus" which are not
understood by limma, so it will not be possible for rbind() to work correctly on your objects.  Is
it possible that you meant to follow the limma User's Guide advice on setting the control status
of each probe (e.g., on page 18)

  RG$genes$Status <- controlStatus(spottypes, RG)

but have missed out the second $-sign?  If you did, that would produce the extra component in the
data objects that we see above.

You can remove the non-standard components by

   RGAmodel$genesStatus <- NULL
   RGBmodel$genesStatus <- NULL

Do the objects rbind() correctly after removing these components?  If the problem persists, can
you please show the output from

  show(RGAmodel)
  show(RGBmodel)

so we can see more information.

Best wishes
Gordon