Hi all,
In principle the package was designed using the 4x44k arrays data and
I did not check the compatibility with other data comming from other
Agilent one-color arrays. We could use the package to process other
Agilent one-color array data, as long as the data format is the same
as the one created by the Agilent Feature Extraction for the 4x44k
arrays, so that we can create a proper RGList. Aside from this, there
are not any other especial requirements that force us to use
exclusively the 4x44k array data so I guess that the package could be
used exactly in the same way, so if this is so, it might make more
sense to rename the package with any other name, as Francois has
pointed out.
When using other Agilent one-color data, we should pay attention to
the following points:
a) The data have to be generated with the Agilent Feature Extraction
9.1.3.1 (or later version). If we use previous version of AFE it will
make the program to crash. Other data files generated with other
image analysis software cannot be used either. Well, this is mandatory
whether or not the data comes from the 4x44k arrays.
b) If we use AFE software to create the input files from other Agilent
one-color array different from the 4x44k, as long as data formats are
exactly the same as the ones created by AFE for 4x44k arrays, the
package will work as it should. We have to be especially cautious
here, since the input data files have to provide all the information
that is needed in our RGList object. You can have a look at the
read.AgilentFE function to see what information is this, but
basically, the columns that we expect to find in the input data files
are:
b1) signal data:
list(Rf = "gProcessedSignal", Gf = "gMeanSignal",
Rb = "gBGMedianSignal", Gb = "gBGUsed")
b2) Flags and other sort of data:
list(IsFound = "gIsFound", IsWellAboveBG =
"gIsWellAboveBG",
IsSaturated = "gIsSaturated", IsFeatNonUnifOF =
"gIsFeatNonUnifOL",
IsFeatPopnOL = "gIsFeatPopnOL", ChrCoord = "chr_coord")
(all what this variable names stand for can be found in the vignette,
page 4. )
If any of these columns were missing, then we could get oursevelves
into troubles either reading the data files or in later steps, since
the RGList object could contain misleading information.
Besides, if there are non replicated probes on the chip (this can be
checked using the CV.rep.probes function) the summarization step can
obviously be skipped.
p.-
Hi Pedro,
Out of curiosity, is there anything specific to 4x44k arrays in the
package, as opposed to 1x44k or 244k arrays?
If not, it might be worth considering renaming the package to
something
a bit more general, like AgiPreProcess or AgiOneColorPreProcess.
Francois
Pedro López Romero wrote:
> Hi everyone,
>
> Sean is right. Agilent advocates now for the use of one-single color
arrays and we have designed Agi4x44PreProcess for the analysis of this
sort of data. To get the data into R, we use the limma function
read.maimages, but we only get the green signals since these are the
only signals that we have in our data files.
>
> To be more precise, the function read.AgilentFE uses the limma
function read.maimages that creates an RGList object, for example:
>
> dd=read.AgilentFE(targets,makePLOT=FALSE)
>
> dd is the RGList that includes
>
> dd$Rf="gProcessedSignal",
> dd$Gf="gMeanSignal",
> dd$Rb="gBGMedianSignal",
> dd$Gb="gBGUsed"
>
> all of which are green signals. Depending on how you want to process
your data you might select one or another signal from the ones
provided according to different methods of backround correction.
>
> The package cannot be used for the processing of two-single color
arrays.
>
> HTH
>
> Pedro
>
> ________________________________
>
> From: bioconductor-bounces@stat.math.ethz.ch on behalf of Sean Davis
> Sent: Tue 2/3/2009 5:33 PM
> To: Erika Melissari
> Cc: bioconductor@stat.math.ethz.ch
> Subject: Re: [BioC] agi4x44kpreprocess package
>
>
>
> On Tue, Feb 3, 2009 at 10:54 AM, Erika Melissari <
> erika.melissari@bioclinica.unipi.it> wrote:
>
>> Hello Francois,
>>
>> thank you for your help.
>>
>> here is the sessionInfo() output
>>
>> R version 2.8.0 (2008-10-20)
>> i386-pc-mingw32
>>
>> locale:
>> LC_COLLATE=English_United States.1252;LC_CTYPE=English_United
>> States.1252;LC_MONETARY=English_United
>> States.1252;LC_NUMERIC=C;LC_TIME=English_United States.1252
>>
>> attached base packages:
>> [1] splines tools stats graphics grDevices utils
datasets
>> methods
>> [9] base
>>
>> other attached packages:
>> [1] hgug4112a.db_2.2.5 RSQLite_0.7-1 DBI_0.2-4
>> [4] Agi4x44PreProcess_1.2.0 genefilter_1.22.0 survival_2.34-1
>> [7] annotate_1.20.1 xtable_1.5-4
AnnotationDbi_1.4.1
>> [10] limma_2.16.3 Biobase_2.2.1
>>
>> I realized that this package is wrapped around limma, but the
result of
>> read.maimages is not the same.
>> I have checked, comparing the loaded data with raw data, what date
have
>> been loaded by using read.AgilentFe of agi4x44preprocess and,
really, only
>> the green signals are loaded. Obviously, the read.maimages() load
the
>> correct data (I have checked).
>> The problem is that agi4x44preprocess vignette also affirm that the
signals
>> loaded are the green signals...but I do not understand why only
green!
>
>
> For gene expression arrays, Agilent has been advocating single-
channel
> protocols for a couple of years. There are groups for which that is
the
> standard.
>
> Sean
>
>
>
>> I am evaluating this package because it offer some functions for
managing
>> the huge amount of data (processed signals, mean signals and, above
all, the
>> flags) extracted by using Feature extraction software by Agilent. I
would
>> like to use the complicated system of Agilent flags for quality
control of
>> my microarray data and only this package manages to do this.
>> Thank you for any suggestion.
>>
>> Erika
>>
>> ----- Original Message ----- From: "Francois Pepin"
<fpepin@cs.mcgill.ca>
>> To: "Erika Melissari" <erika.melissari@bioclinica.unipi.it>
>> Cc: <bioconductor@stat.math.ethz.ch>
>> Sent: Tuesday, February 03, 2009 16:20 PM
>> Subject: Re: [BioC] agi4x44kpreprocess package
>>
>>
>> Hi Erika,
>>> I have never used that package, but it seems to be mostly a
wrapper
>>> around limma for those steps. What happens if you use
>>> read.maimages(files, source='agilent') directly?
>>>
>>> You might also want to make sure that you have red values if you
look
>>> directly in your raw text file.
>>>
>>> And please post the output from sessionInfo(). This will make it
easier
>>> for people to understand which type of installation you have, and
>>> generally makes it easier to catch if the problem.
>>>
>>> Francois
>>>
>>> Erika Melissari wrote:
>>>
>>>> Hello all,
>>>>
>>>> does anyone know Agi4x44kPreProcess package?
>>>> I'm exploring it to perform pre-process steps on human 4x44k
Agilent
>>>> microarrays, but I have a question or perhaps a doubt: why does
this package
>>>> load only the green signals (processed, mean, ect.) and not the
red signals?
>>>> Is it an error of this package or a characteristic?
>>>> Thank you so much for any indication
>>>>
>>>> Erika
>>>>
>>>> [[alternative HTML version deleted]]
>>>>
>>>> _______________________________________________
>>>> Bioconductor mailing list
>>>> Bioconductor@stat.math.ethz.ch
>>>>
https://stat.ethz.ch/mailman/listinfo/bioconductor
>>>> Search the archives:
>>>>
http://news.gmane.org/gmane.science.biology.informatics.conductor
>>>>
>>>
>>
>> -------------------------------------------------------------------
-------------
>>
>>
>>
>> Nessun virus nel messaggio in arrivo.
>> Controllato da AVG - www.avg.com
>> Versione: 8.0.233 / Database dei virus: 270.10.17/1932 - Data di
rilascio:
>> 02/02/09 07:51:00
>>
>> _______________________________________________
>> Bioconductor mailing list
>> Bioconductor@stat.math.ethz.ch
>>
https://stat.ethz.ch/mailman/listinfo/bioconductor
>> Search the archives:
>>
http://news.gmane.org/gmane.science.biology.informatics.conductor
>>
>
> [[alternative HTML version deleted]]
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor@stat.math.ethz.ch
>
https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives:
http://news.gmane.org/gmane.science.biology.informatics.conductor
>
>
>
> [[alternative HTML version deleted]]
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor@stat.math.ethz.ch
>
https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives:
http://news.gmane.org/gmane.science.biology.informatics.conductor
[[alternative HTML version deleted]]