[BioC] arrayQualityMetrics package - bugs and errors

[Wolfgang Huber]

Dear Vladimir

thank you for the below message, and sorry for the delayed reply. Your 
problem is related to this thread:
https://stat.ethz.ch/pipermail/bioconductor/2012-March/043888.html

I.e. on the short term, to use 'MAList' objects I recommend that you 
convert them yourself with
	as( MA, "ExpressionSet")

On my to do list:
- change arrayQualityMetrics so that it accepts 'MAList' and 
automatically converts them into 'ExpressionSet'.
- make the automatic conversion more tolerant of the different ways that 
MALists are instantiated by limma.

	Best wishes
	Wolfgang

Feb/16/12 1:52 PM, Vladimir Krasikov scripsit::
> Dear Wolfgang
>
> Thanks a lot for your advises.
> Bellow I provide error messages for my case
>
> On Wed, 15 Feb 2012 19:42:27 +0100, Wolfgang Huber <whuber at embl.de> wrote:
>
>> Dear Vladimir
>>
>> Thank you for the feedback and the detailed descriptions.
>>
>> To remove the reason for the warning from KernSmooth::bkde2D (which is
>> called by the MA-plots section), please use the argument
>> 'do.logtransform=TRUE' for the function 'arrayQualityMetrics'. Does
>> that resolve your problem?
>
> Thanks, log-transformation indeed resolves warnings.
> And also this step brought some meaning to the Report.
>
>> You say "it doesn't work with MAlist objects at all". Can you please
>> provide the exact error message and a reproducible example?
>
> arrayQualityMetrics(expressionset = MA, outdir = "D:/R2/Out/Report",
> force = TRUE, do.logtransform=TRUE)
>
> The directory 'D:/R2/Out/Report' has been created.
> Error in as(expressionset, "NChannelSet") :
> no method or default for coercing "MAList" to "NChannelSet"
> Error in prepdata(expressionset, intgroup = intgroup, do.logtransform =
> do.logtransform) :
> Argument 'expressionset' is of class 'MAList', and its automatic
> conversion into 'NChannelSet' failed. Please try to convert it manually.
>
>> Interoperability of 'RGList' and 'MAList' objects with the rest of
>> Bioconductor can be challenging. arrayQualityMetrics uses the
>> 'convert' package to convert them into 'NChannelSet', and then works
>> from that. If you have problems with the phenoData, please send us a
>> reproducible example so we can try to update the conversion code.
>
> I guess pData method is not defined for the RGList (and the MAList) object
> and arrayQualityMetrics first tries to set phenoData for the RGList and
> then to convert it to the NChannelSet
>
> pData(RG)$Cy3 <- targets$Cy3
> Error in function (classes, fdef, mtable) :
> unable to find an inherited method for function "pData", for signature
> "RGList"
>
>> As a short term solution to your problem, do the conversion yourself with
>>
>> library("convert")
>> x = as(RG, "NChannelSet")
>>
>> and then set the phenoData of x as you wish.
>
> And above solution works perfect and add some colors to otherwise nice
> report.
>
> Thanks a lot
> Vladimir
>
> and as recommended sessionInfo:
>> sessionInfo()
> R version 2.14.1 (2011-12-22)
> Platform: i386-pc-mingw32/i386 (32-bit)
>
> locale:
> [1] LC_COLLATE=Dutch_Netherlands.1252 LC_CTYPE=Dutch_Netherlands.1252
> LC_MONETARY=Dutch_Netherlands.1252
> [4] LC_NUMERIC=C LC_TIME=Dutch_Netherlands.1252
>
> attached base packages:
> [1] stats graphics grDevices utils datasets methods base
>
> other attached packages:
> [1] convert_1.30.0 marray_1.32.0 arrayQualityMetrics_3.10.0 Biobase_2.14.0
> [5] limma_3.10.2 BiocInstaller_1.2.1
>
> loaded via a namespace (and not attached):
> [1] affy_1.32.1 affyio_1.22.0 affyPLM_1.30.0 annotate_1.32.1
> [5] AnnotationDbi_1.16.16 beadarray_2.4.1 Biostrings_2.22.0 Cairo_1.5-1
> [9] cluster_1.14.2 DBI_0.2-5 genefilter_1.36.0 grid_2.14.1
> [13] Hmisc_3.9-2 hwriter_1.3 IRanges_1.12.6 KernSmooth_2.23-7
> [17] lattice_0.20-0 latticeExtra_0.6-19 preprocessCore_1.16.0
> RColorBrewer_1.0-5
> [21] RSQLite_0.11.1 setRNG_2009.11-1 splines_2.14.1 survival_2.36-12
> [25] SVGAnnotation_0.9-0 tools_2.14.1 vsn_3.22.0 XML_3.9-4.1
> [29] xtable_1.7-0 zlibbioc_1.0.0
>
>
>
>
>
>>
>> best wishes
>> Wolfgang
>>
>> Vladimir Krasikov scripsit 02/15/2012 01:44 PM:
>>> Dear list
>>>
>>> While trying to analyze my data with arrayQualityMetrics (thanks to
>>> Axel Klenk for the advise to use it) first on PC and now on MAC.
>>> I realize that it is can't run at all on MAC, while on PC it does work,
>>> although not as described (or at least not all features described in the
>>> vignette are functional)
>>>
>>> 1. MAC story first:
>>> It cann't be loaded without affyPLM (so I guess arrayQualityMetrics
>>> DEPENDS on affyPLM,
>>> and probably on all other packages which are listed in imports field)
>>>
>>>> library(arrayQualityMetrics)
>>> Creating a generic function for ‘boxplot’ from package ‘graphics’ in
>>> package ‘affyPLM’
>>> Creating a generic function for ‘hist’ from package ‘graphics’ in
>>> package ‘affyPLM’
>>> Error in as.environment(pos) :
>>> no item called "newtable" on the search list
>>> In addition: Warning message:
>>> In objects(newtable, all.names = TRUE) :
>>> ‘newtable’ converted to character string
>>> Error: package/namespace load failed for ‘arrayQualityMetrics’
>>>
>>> MAC sessionInfo:
>>>> sessionInfo()
>>> R version 2.14.1 (2011-12-22)
>>> Platform: i386-apple-darwin9.8.0/i386 (32-bit)
>>>
>>> locale:
>>> [1] en_GB.UTF-8/en_GB.UTF-8/en_GB.UTF-8/C/en_GB.UTF-8/en_GB.UTF-8
>>>
>>> attached base packages:
>>> [1] stats graphics grDevices utils datasets methods base
>>>
>>> other attached packages:
>>> [1] vsn_3.22.0 Biobase_2.14.0 limma_3.10.2
>>>
>>> loaded via a namespace (and not attached):
>>> [1] affy_1.32.1 affyio_1.22.0 BiocInstaller_1.2.1 grid_2.14.1
>>> lattice_0.20-0
>>> [6] preprocessCore_1.16.0 tools_2.14.1 zlibbioc_1.0.0
>>>>
>>>
>>> #################
>>> 2. PC sorry next:
>>>
>>> Installation also required to install all packages one-by-one from the
>>> "imports"
>>>
>>> I think it doesn't work properly with objects generated by limma (or at
>>> least in my case):
>>>
>>> I did generic import of 61x44K Agilent two-colours microarrays:
>>>> RG <- read.maimages(files....)
>>> and background correction
>>>> RG.b <- backgroundCorrect(RG.raw, method="minimum", offset=50)
>>>> arrayQualityMetrics(expressionset = RG.b, outdir =
>>>> "D:/R2/Out/Report.61.Rq", force = TRUE)
>>> produces 18 warnings of this type:
>>>> warnings()
>>> 18: In KernSmooth::bkde2D(x, gridsize = nbin, bandwidth = bandwidth) :
>>> Binning grid too coarse for current (small) bandwidth: consider
>>> increasing 'gridsize'
>>>
>>> If to create MA from RG with: >MA <- MA.RG(RG.b)
>>>
>>> It doesn't work with MAlist objects at all,
>>> and it is not able to create phenoData for RGlist also (which is quite a
>>> pity)
>>> [I can't reproduce it on MAC, as I can't start the package at all,
>>> I will update this error tomorrow from PC]
>>>
>>> The rest of the report is generated well for the RG object
>>>
>>> PC sessionInfo:
>>>> sessionInfo()
>>> R version 2.14.1 (2011-12-22)
>>> Platform: i386-pc-mingw32/i386 (32-bit)
>>>
>>> locale:
>>> [1] LC_COLLATE=Dutch_Netherlands.1252 LC_CTYPE=Dutch_Netherlands.1252
>>> [3] LC_MONETARY=Dutch_Netherlands.1252 LC_NUMERIC=C
>>> [5] LC_TIME=Dutch_Netherlands.1252
>>>
>>> attached base packages:
>>> [1] stats graphics grDevices utils datasets methods base
>>>
>>> other attached packages:
>>> [1] arrayQualityMetrics_3.10.0 Biobase_2.14.0 limma_3.10.2
>>>
>>> Dear authors of arrayQualityMetrics, may you comment?
>>>
>>> With kind regards
>>> Vladimir
>>>
>>>
>>>
>>>
>>>
>>
>>
>
>


-- 
Best wishes
	Wolfgang

Wolfgang Huber
EMBL
http://www.embl.de/research/units/genome_biology/huber