[BioC] Biobase - Error: ExpressionSet
Laurent Gatto
lg390 at cam.ac.uk
Thu Oct 17 22:09:18 CEST 2013
lisa.johnson.cohen at gmail.com writes:
> Hi Laurent, Thank you for your message. It helped tremendously. I converted
> my data to a matrix and also had to fiddle around a bit get the row and
> column names to match up. But, I successfully made the ExpressionSet! Thank
> you so much.
Great!
> sponge_ExpressionSet<-new("ExpressionSet",exprs=sponge_data,phenoData=pd,experimentData=experimentData,featureData=an)
>
> Do you by any chance work with custom microarrays? We use custom Agilent
> microarrays. Do you have any suggetions for quality assessment?
The best place to look for microarray quality is probably the
arrayQualityMetrics package [1]. I have never tried it with an
ExpressionSet built from a custom Agilent array though.
Best wishes,
Laurent
[1] http://www.bioconductor.org/packages/release/bioc/html/arrayQualityMetrics.html
> Thank you again!
> Lisa
>
> ---
> Lisa Cohen, M.Sc.
> Marine Ecosystem Health,
> Harbor Branch Oceanographic Institute
> at Florida Atlantic University
> 5600 N. US Hwy 1
> Fort Pierce, FL 34946
> Phone: (772) 242-2396
> Fax: (772) 468-0757
> Email:lcohen49 at hboi.fau.edu
> Alt. Email: lisa.johnson.cohen at gmail.com
>
> <lcohen49 at hboi.fau.edu>
>
>
>
>
>
> On Wed, Oct 16, 2013 at 4:44 PM, Laurent Gatto <lg390 at cam.ac.uk> wrote:
>
>>
>> Dear Lisa,
>>
>> > I'm using the Biobase package to build an ExpressionSet from scratch
>> using
>> > data from a custom 8x15k Agilent microarray. I'm getting this error and
>> > can't quite figure it out. Does anyone know how *phenoData* can extend
>> the *
>> > annotation*? I would like to connect *phenoData* to *exprs* with the Chip
>> > and Block numbers, somehow, but I don't know how to do this. Any help
>> would
>> > be greatly appreciated!
>> >
>> >>
>> sponge_ExpressionSet<-new("ExpressionSet",exprs=sponge_data,phenoData=samples_matrix,experimentData=experimentData,annotation=annotation_matrix)Error:
>> ExpressionSet 'phenoData' is class 'matrix' but should be or extend
>> 'AnnotatedDataFrame'
>>
>> The message tells you that the content of the phenoData slot must be an
>> 'AnnotateDataFrame'. Once you have it as a data.frame, you can do the
>> following
>>
>> pd <- new("AnnotatedDataFrame", data = pddataframe)
>>
>> and then pass phenoData = pd in the constructor.
>>
>> This also holds for the featureData slot.
>>
>> The expression data, on the other hand must be a matrix (not a
>> data.frame).
>>
>> Hope this helps.
>>
>> Best wishes,
>>
>> Laurent
>>
>> >
>> > Details:
>> >
>> > These are the instructions that I am following:
>> >
>> http://www.bioconductor.org/help/course-materials/2006/biocintro_oct/labs/ExpressionSet/ExpressionSetIntro-lab.pdf
>> >
>> http://www.bioconductor.org/packages/2.13/bioc/vignettes/Biobase/inst/doc/ExpressionSetIntroduction.pdf
>> >
>> > 1. *exprs* data (sponge_data) were extracted from GenePix *.gpr files and
>> > are in a 15744 x 72 double matrix with RefNumber as rownames:
>> >
>> >> head(sponge_data)
>> >
>> > Block 1 Chip 1 Block 2 Chip 1 Block 3 Chip 1 Block 4 Chip 1 Block
>> 5 Chip 1
>> > 15744 2 23 6 9
>> 3
>> > 15580 9 16 11 11
>> 6
>> > 15416 1 0 0 0
>> 1
>> > 15252 1 0 0 -1
>> 0
>> > 15088 3 2 0 1
>> 0
>> > 14924 0 -1 0 0
>> 1
>> >
>> >
>> > 2. *phenoData* (samples_matrix) is a 72 x 8 character matrix with
>> > experimental design variables for each chip and block:
>> >
>> >> head(samples_matrix) Chip Number File Name Hyb
>> Date Tank Treatment Colony Raceway
>> > [1,] "1" "253295110067_2012-08-01.txt" "8/1/2012" "C3" "UC"
>> > "P" "B"
>> > [2,] "1" "253295110067_2012-08-01.txt" "8/1/2012" "B2" "UD"
>> > "P" "A"
>> > [3,] "1" "253295110067_2012-08-01.txt" "8/1/2012" "B2" "UD"
>> > "O" "A"
>> > [4,] "1" "253295110067_2012-08-01.txt" "8/1/2012" "D3" "OD"
>> > "BL" "A"
>> > [5,] "1" "253295110067_2012-08-01.txt" "8/1/2012" "D1" "OD"
>> > "O" "B"
>> > [6,] "1" "253295110067_2012-08-01.txt" "8/1/2012" "C3" "UC"
>> > "O" "B"
>> >
>> >> all(rownames(samples_matrix)==colnames(sponge_data))[1] TRUE
>> >
>> >
>> > 3. *annotation* is (annotation_matrix) is a 15744 x 23 character matrix
>> > with microarray annotation info, including RefNumber:
>> >
>> >> head(annotation_matrix) Column Row RefNumber Name Append
>> Name
>> > [1,] "96" "164" " 1" "GE_BrightCorner" "GE_BrightCorner"
>> > [2,] "96" "162" " 2" "DarkCorner" "DarkCorner"
>> > [3,] "96" "160" " 3" "DarkCorner" "DarkCorner"
>> > [4,] "96" "158" " 4" "222803147_EZ031535.1"
>> > "222803147_EZ031535.1_16-465"
>> > [5,] "96" "156" " 5" "222546129_EZ002319.1"
>> > "222546129_EZ002319.1_1845-2030"
>> > [6,] "96" "154" " 6" "222801996_EZ030384.1"
>> > "222801996_EZ030384.1_31-2558"
>> >
>> >> all(rownames(annotation_matrix)==rownames(sponge_data))[1] TRUE
>> >
>> >
>> > 4. experimentData (experimentData): metadata info
>> >
>> > But, when I go to finally assemble the ExpressionSet, the error below
>> > occurs. I think the problem is in connecting *phenoData* to *annotation*,
>> > but I'm not sure how to do this. I would like to connect the *exprs* to *
>> > annotation*, to connect Chip and Block numbers, but I don't see how to do
>> > this. Any help would be greatly appreciated!
>> >
>> >>
>> sponge_ExpressionSet<-new("ExpressionSet",exprs=sponge_data,phenoData=samples_matrix,experimentData=experimentData,annotation=annotation_matrix)Error:
>> ExpressionSet 'phenoData' is class 'matrix' but should be or extend
>> 'AnnotatedDataFrame'
>> >
>> >
>> >
>> > Thank you!
>> > Lisa
>> >
>> > [[alternative HTML version deleted]]
>> >
>> > _______________________________________________
>> > 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
>>
>>
More information about the Bioconductor
mailing list