[BioC] affyxparser package question
Benilton Carvalho
bcarvalh at jhsph.edu
Thu Aug 9 00:43:46 CEST 2007
What happens is pretty the same effect of what you get when you do
as.numeric() on a matrix...
There is a one to one mapping between X and Y coordinates (given that
you the dimensions of the chip) to an index... This is the nature of
the xy2i function used a lot on the affy package and everything else
on BioConductor. And that's how we get the vector.
The basic idea behind the CEL file (over-simplifying here) is that it
contains 3 columns that we use and I could mimic it with something like:
ints <- rnorm(80)
fake.cel <- expand.grid(X=1:10, Y=1:8)
fake.cel <- cbind(fake.cel, intensity=ints)
if you have a few minutes, you'll see that there is an association
between the X and Y columns with the rownumber... this rownumber is
the "index" I referred to above and is the "fid" column in the
annotation.
b
On Aug 8, 2007, at 6:33 PM, Tae-Hoon Chung wrote:
> Hi, Benilton;
>
> When I tried the 'low-level' method of using readCelUnits(), I got
> a list with individual probe intensities. However, when I tried
> read.celfiles() from 'oligo' package, I got an object of
> 'SnpFeatureSet' class whose measurement values were retrieved
> through exprs() method. However, the matrix retrieved by this
> method contained one vector for each sample and it was hard to
> figure out how the low-level intensity values were transformed into
> a single vector. I am curious if the low-level intensity values are
> still preserved in the SnpFeatureSet object or more or less
> summarized in it.
>
> Tae-Hoon Chung
>
> Post-Doctoral Researcher
> Computational Biology Division, TGEN
> 445 N 5th St. Phoenix, AZ 85004 USA
> O: 1-602-343-8724
> F: 1-602-343-8840
>
>
> On Aug 8, 2007, at 3:24 PM, Benilton Carvalho wrote:
>
>> I don't mean that it's going to be very hard.... I mean that it
>> might give you some work and someone else already figured that out
>> for you and you could have been using your time for something else.
>>
>> All you need is something like:
>>
>> library(pd.mapping50k.xba240)
>> dbGetQuery(db(pd.mapping50k.xba240), "SELECT * from pmfeature
>> LIMIT 5")
>>
>> b
>>
>> On Aug 8, 2007, at 6:15 PM, Tae-Hoon Chung wrote:
>>
>>> Hi, Benilton;
>>>
>>> Does this mean that it's going to be tough to tell which ones are
>>> for forward/backward strand or for central/shifted probe without
>>> involving annotation package?
>>>
>>> Tae-Hoon Chung
>>>
>>> Post-Doctoral Researcher
>>> Computational Biology Division, TGEN
>>> 445 N 5th St. Phoenix, AZ 85004 USA
>>> O: 1-602-343-8724
>>> F: 1-602-343-8840
>>>
>>>
>>> On Aug 8, 2007, at 3:08 PM, Benilton Carvalho wrote:
>>>
>>>> any specific reason to not use
>>>>
>>>> library(oligo)
>>>> x = read.celfiles(list.celfiles())
>>>>
>>>> and check the annotation packages?
>>>>
>>>> b
>>>>
>>>> On Aug 8, 2007, at 6:02 PM, Tae-Hoon Chung wrote:
>>>>
>>>>> Hi;
>>>>>
>>>>> When one uses readCelUnits() to read SNP chip cel files, how
>>>>> one can
>>>>> tell which values are for forward strand or for backward strand
>>>>> and
>>>>> which values are from the non-shifted probes or from the shifted
>>>>> probes? For instance, in the following code chunk, which values
>>>>> are
>>>>> from forward/backward strand and from central/shifted probes?
>>>>>
>>>>> library("hapmap100kxba", lib="~/Library/R64")
>>>>> library(affxparser, lib="~/Library/R64")
>>>>>
>>>>> pth <- system.file('celFiles', package='hapmap100kxba')
>>>>> files <- list.files(path=pth, full.names=T)
>>>>>
>>>>> chip.type <- readCelHeader(files[1])$chiptype ##
>>>>> Mapping50K_Xba240
>>>>> cels <- readCelUnits(files[1], cdf='~/Project/ProbeAnnot/
>>>>> Mapping50K_Xba240.cdf', stratifyBy='pm', addDimnames=T)
>>>>> length(cels) ## 59015
>>>>> labs.test <- names(cels)[100:120]
>>>>> cels[[labs.test[1]]]
>>>>> ## $A
>>>>> ## $A$intensities
>>>>> ## [1] 7563 8050 9531 9292 11261
>>>>> ##
>>>>> ## $G
>>>>> ## $G$intensities
>>>>> ## [1] 6540 7639 9027 10512 11381
>>>>> ##
>>>>> ## $A
>>>>> ## $A$intensities
>>>>> ## [1] 4036 4144 3858 5170 3975
>>>>> ##
>>>>> ## $G
>>>>> ## $G$intensities
>>>>> ## [1] 4425 4291 3682 5912 5208
>>>>>
>>>>>
>>>>> Tae-Hoon Chung
>>>>>
>>>>> Post-Doctoral Researcher
>>>>> Computational Biology Division, TGEN
>>>>> 445 N 5th St. Phoenix, AZ 85004 USA
>>>>> O: 1-602-343-8724
>>>>> F: 1-602-343-8840
>>>>>
>>>>>
>>>>>
>>>>> [[alternative HTML version deleted]]
>>>>>
>>>>> _______________________________________________
>>>>> Bioconductor mailing list
>>>>> Bioconductor at stat.math.ethz.ch
>>>>> 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