[BioC] Unique probes on all human affy chips

Laurent Gautier lgautier at altern.org
Mon Sep 20 06:33:53 CEST 2004


mypacks <- c("hgu133aprobe", "hgu95av2probe") ## put all the cdfenv 

                                               ## packs you want here

allmyprobes <- c()

for (i in seq(along=mypacks)) {
   library(mypacks[i], character.only=TRUE)
   allmyprobes <- unique(c(get(mypacks[i])$sequence, allmyprobes))

   ## memory saving...
   tmp <- paste("package", mypacks[i], sep=":")
   detach(pos=match(tmp, search()))
}

## unique probe sequences are in 'allprobes'...
## ... and shouldn't take 20 hours for 13 chips...

If you feel like leaving C++ aside:
- To map probes to reference sequences the package 'matchprobes'
can be useful.
- To manipulate mappings, the package 'altcdfenvs' might have functions 
of interest as well.



L.




S Peri wrote:
> Hi Laurent, 
>  I am interested in getting individual probes. I want
> to export them into my postgres table and map link
> them to sequences. 
> Thanks
> Peri. 



> --- Laurent Gautier <lgautier at altern.org> wrote:
> 
> 
>>Peri,
>>
>>
>>I am not certain of what you want: are they the
>>probe sets
>>or the individual probes ?
>>If you want probe sets, I think that you could do it
>>in R. The execution time would be few seconds (on
>>reasonably recent 
>>computer...)
>>
>>Example:
>>
>>mycdfs <- c("hgu133acdf", "hgu95av2cdf") ## put all
>>the cdfenv packs you
>>                                         ## want
>>here
>>
>>allmyids <- c()
>>
>>for (i in seq(along=mycdf)) {
>>   library(mycdfs[i], character.only=TRUE)
>>   allmyids <- unique(c(ls(get(mycdfs[i])),
>>allmyids))
>>
>>   ## memory saving...
>>   tmp <- paste("package", mycdfs[i], sep=":")
>>   detach(pos=match(tmp, search()))
>>}
>>
>>## unique probe set ids are in 'allmyids'...
>>
>>
>>If you want the probes, a variation of the code
>>above should make
>>it.
>>
>>
>>Hoping it helps,
>>
>>
>>
>>L.
>>
>>
>>
>>S Peri wrote:
>>
>>>Dear group, 
>>> Is there any place where I can get all the unique
>>>probe ids for all the Affy human chips (~13
>>
>>chips). 
>>
>>>I am trying to get the unique probes (no
>>
>>duplicates).
>>
>>>It turned out to be very computing intensive
>>
>>problem. 
>>
>>>I took all the probes from all 13 chips and made a
>>>program that writes the the unique id (if there
>>
>>are
>>
>>>duplicates, for e.g. 64474_g_at is there on
>>
>>HG-U95C,
>>
>>>HGU133, HGU133A2, and 133_plus2. In this case my
>>>program will write 64474_g_at once in my output).
>>>Using c++ code it is running for the last 20 hrs.
>>
>>I
>>
>>>made sure there are no bad loops that would put me
>>
>>in
>>
>>>infinite loop situation. It would be nice to have
>>
>>all
>>
>>>the uniqe ids in some place where i can use them
>>>directly for my annotation purposes. 
>>>
>>>Thanks
>>>Peri.
>>>
>>>_______________________________________________
>>>Bioconductor mailing list
>>>Bioconductor at stat.math.ethz.ch
>>>https://stat.ethz.ch/mailman/listinfo/bioconductor
>>>
>>
>>
> 
> 
> 
> 		
> _______________________________
> Do you Yahoo!?
> Declare Yourself - Register online to vote today!
> http://vote.yahoo.com
>



More information about the Bioconductor mailing list