[BioC] Problems using arrayQualityMetrics

[Wolfgang Huber]

Dear Syed

Mar/23/12 11:57 AM, Syed Haider scripsit::
> Thanks Wolfgang for your detailed reply.
> The machine which i tried this on had:
>
> arrayQualityMetrics v3.2.4 (installed from source)
> R v2.12.2
> System wide Cairo v1.2.4
> R Cairo v1.5-1

All of these are old and outdated.

> I guess Ros et al have tried this on the latest versions too.

Ros used a recent R and Bioconductor, but still a much too old cairo 
graphics library.

	Best wishes
	Wolfgang

>
> Thanks,
> Syed
>
>
>
> On 23/03/2012 08:12, Wolfgang Huber wrote:
>> Dear Syed
>>
>> thanks! With what version of 'arrayQualityMetrics' did you experience
>> this? As far as I can tell, the function 'annotateSvgMatplot' was last
>> used in Bioconductor release 2.7 from autumn 2010.
>>
>> To use this package, please, please use a *current version of R,
>> Bioconductor, and the cairo graphics* library.
>>
>> Please do not expect that any old combination of these components will
>> work together. Also, you would be missing out on the cool new features.
>>
>> I recommend using
>> http://www.bioconductor.org/packages/devel/bioc/html/arrayQualityMetrics.html,
>>
>> which is close to release.
>>
>> PS - OKish looking plots are one thing, but SVGAnnotation postprocesses
>> the SVG code for the plots and requires that that code is quite
>> precisely as expected.
>>
>> Enjoy,
>> Wolfgang
>>
>> Mar/23/12 1:04 AM, Syed Haider scripsit::
>>> Dear Wolfgang,
>>>
>>> Ros's computer definitely needs an upgrade. However, looking breifly
>>> into this, here is the offending line of code from
>>> arrayQualityMetrics::aqm.report.qm(...)
>>>
>>> size = annotateSvgMatplot(svgtemp, nameimg, annotationInfo=qm$svg)
>>>
>>> The above fails while generating pca.svg and rnadeg.svg. Interestingly
>>> density.svg comes out *okish*. The density plot looks correct but
>>> without axis-labels and legends. All other figures are fine too,
>>> including the pdf equivalent of all svgs.
>>>
>>>
>>> Best,
>>> Syed
>>>
>>>
>>> On 22/03/2012 23:39, Wolfgang Huber wrote:
>>>> Dear Rosalind
>>>>
>>>> unfortunately, the fact that the package manager on your Redhat Linux
>>>> does not provide you with a newer version does not mean that there is
>>>> none. The Linux distribution that you are using seems to be very much
>>>> behind in time: cairo 1.2.4 is from 2006, see http://cairographics.org/
>>>>
>>>> Two options:
>>>> - use a system that is better kept up-to-date
>>>> - remove the cairo that your package manager gives you, and install
>>>> libcairo >=1.10.2 'manually' by downloading it from the above URL.
>>>>
>>>> Btw, I would also also prefer if arrayQualityMetrics (i.e., really the
>>>> SVGAnnotation package) was backwards compatible with any old version of
>>>> cairo, but that is not a realistic expectation (neither for our
>>>> software, nor for many other products, incl. commercial).
>>>>
>>>> Best wishes
>>>> Wolfgang
>>>>
>>>> Mar/22/12 12:38 PM, Rosalind Cutts scripsit::
>>>>> Thanks very much for the reply, I have tried this with a clean R and
>>>>> bioconductor installation and get a similar error.
>>>>>
>>>>> ArrayQualityMetrics works ok, it just fails to produce some of the
>>>>> graphical output.
>>>>>
>>>>> Cairo version:
>>>>> pkg-config --modversion cairo
>>>>> 1.2.4
>>>>>
>>>>> When I tried updating Cairo on the Redhat linux machine I am using as
>>>>> suggested by your link there were no updates to load.
>>>>>
>>>>> Is there anything else that you could suggest I could try.
>>>>>
>>>>> Thanks again for your help
>>>>>
>>>>> Ros
>>>>
>>>>
>>>> Wolfgang Huber
>>>> EMBL
>>>> http://www.embl.de/research/units/genome_biology/huber
>>>>
>>>> _______________________________________________
>>>> Bioconductor mailing list
>>>> Bioconductor at r-project.org
>>>> https://stat.ethz.ch/mailman/listinfo/bioconductor
>>>> Search the archives:
>>>> http://news.gmane.org/gmane.science.biology.informatics.conductor
>>
>>


-- 
Best wishes
	Wolfgang

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