[BioC] gcrma problem caused by using updated new CDF and probesequence?

yzhang at vbi.vt.edu yzhang at vbi.vt.edu
Fri Sep 14 07:34:59 CEST 2007


Dear Zhijin:
I used R 2.6alpha devel version to generate cdf and probe packages.
But when I run GCRMA, I got the following error message:

Adjusting for optical effect..Done.
Computing affinities.Error in tmp.exprs[pmIndex[subIndex]] = apm :
  NAs are not allowed in subscripted assignments

I don't have problem to do GCRMA using my original cdf and probesequence
files.

I put all of original and updated CDF and probesequence files, and
packages which generated by R 2.6 on the following url:
http://ci.vbi.vt.edu/yan/newcdf/ehis_index.html


Thank you very much for your time.

Best
yan








On Wed, September 12, 2007 4:36 pm, Zhijin Wu wrote:
> Hi Yan,
> I will need the source file for a probepackage so that I can install the
> probepacakge, in order to test out the function compute.affinites for you.
> Could you contact the maintainer for the matchprobe package to
> solve that problem first?
>
> Zhijin
> Yan Zhang wrote:
>
>> Dear Dr. Wu:
>> There is no NA in my probesequence file.
>> I downloaded newest version matchprobes
>> package(1.8.1), I got the following error message:
>>
>> makeProbePackage("ehis1a520285f",version="1.0",species="ehis",maintaine
>> r="yanzhang<yzhang at vbi.vt.edu>",build=FALSE,
>>
>> check=FALSE, force=True) Importing the data.
>> Error in rep(NA, max(pm1, mm1, pm2, mm2)) :
>> invalid 'times' argument In addition: Warning messages:
>> 1: NAs introduced by coercion in: as.integer.default(dat[[2]])
>> 2: NAs introduced by coercion in: as.integer.default(dat[[3]])
>> 3: NAs introduced by coercion in: as.integer.default(dat[[4]])
>>
>>
>> Using newest version, I even could not generate probe package.
>>
>>
>> I put cdf file and probesequence file at this url:
>>
>>
>> http://ci.vbi.vt.edu/yan/newcdf/ehis_index.html
>>
>>
>> Thank you very much for your time.
>> best yan
>>
>>
>>
>> Zhijin Wu wrote:
>>
>>
>>> I am not sure what's causing the problem, since we actually allow NAs
>>>  in the sequences. I wonder if this has anything to do with your
>>> trimmed probe package. Can you check first by loading the
>>> probepackage, and use
>>>
>>> myseq <- get("probepackageName")
>>>
>>> and see how many NAs do you get in table(is.na(myseq$seq))?
>>>
>>> You may also send me source file of your probepackage. I can't
>>> promise to test it immediately but will try my best.
>>>
>>>
>>> yzhang at vbi.vt.edu wrote:
>>>
>>>> Dear Dr. Wu:
>>>>
>>>>
>>>> Thank you very much for your fast response. The R version is 2.5
>>>> and I am
>>>> using bioconductor 2.0. I think may be I did not descripte my
>>>> question clearly. My new CDF file is an updated CDF file. Probeset
>>>> name changed. Some probe reassign to new probeset, some are deleted.
>>>> My new probeequence file is
>>>> also a trimmed file compare to old probesequence file. But cel file
>>>> doesn't change(this is an customer made affymetrix chip). The
>>>> reason why I have the new updated CDF and probesequence file is
>>>> because that TIGR reannotation this bacteria genome and correctted a
>>>> lot of mistake. I believe that this kind of problem(reannoation
>>>> genome and have new CDF and probesequence file) will become more and
>>>> more popular in the near future. For using RMA, I don't have problem
>>>> to using new CDF file. But, If I want to using GCRMA, I will receive
>>>> Error message:
>>>>
>>>>
>>>> Error in compute.affinities(cdfName(object), verbose = verbose) :
>>>> NAs are not allowed in subscripted assignments
>>>>
>>>>
>>>> Could you help me to figure out what cause this problem? How could
>>>> I
>>>> solve it?
>>>>
>>>>
>>>> Best
>>>> yan
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> On Fri, September 7, 2007 1:17 pm, Zhijin Wu wrote:
>>>>
>>>>
>>>>> which versions are you using( your affy, gcrma and cdf packages)?
>>>>> I
>>>>> don't have a problem with the hgu95a or hgu133a platforms.
>>>>> yzhang at vbi.vt.edu wrote:
>>>>>
>>>>>
>>>>>> Dear Dr. Zhijin Wu:
>>>>>>
>>>>>>
>>>>>>
>>>>>> This is Yan Zhang from VBI.
>>>>>> From the retrived message of bioconductor mailing list,
>>>>>> https://stat.ethz.ch/pipermail/bioconductor/2005-April/008338.ht
>>>>>> ml I noticed that you and James helped Wenqiong Chen to solve
>>>>>> the gcrma problem cause by using new CDF and probesequence
>>>>>> files. Could you recall how you guys finally solve that problem?
>>>>>>
>>>>>> My problem is I am trying to use the new CDF  and probesequence
>>>>>> files to do gcrma normalization for using gcrma function in R. I
>>>>>> can use new cdf file to generate new cdf envirment and do RMA. I
>>>>>> can also generate probe package and install it without any
>>>>>> problem. But when I run gcrma, I got the following error message:
>>>>>>
>>>>>>
>>>>>>
>>>>>> Error in compute.affinities(cdfName(object), verbose = verbose)
>>>>>> :
>>>>>> NAs are not allowed in subscripted assignments
>>>>>>
>>>>>>
>>>>>>
>>>>>> If I switch to old CDF and probesequence file, I don't have
>>>>>> this problem.
>>>>>>
>>>>>> For the new sequence file, I did following modification:
>>>>>> 1. remove probe type coloumn(if I don't remove it, I could not
>>>>>> make probe package) 2. added affymetrix control probe sequence
>>>>>> from old probe squence file.(doesn't improve, still get the same
>>>>>> error message).
>>>>>>
>>>>>>
>>>>>> I put the commands which I used in error messasge file. I also
>>>>>> put new and old cdf, and probesequence files on the web. You
>>>>>> could download them from url:
>>>>>> http://ci.vbi.vt.edu/yan/newcdf/ehis_index.html
>>>>>>
>>>>>>
>>>>>>
>>>>>> Could you help me to figure out this issue?
>>>>>>
>>>>>>
>>>>>>
>>>>>> best yan
>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>> --
>>>>> -------------------------------------------
>>>>> Zhijin (Jean) Wu
>>>>> Assistant Professor of Biostatistics
>>>>> Brown University, Box G-S121
>>>>> Providence, RI  02912
>>>>>
>>>>>
>>>>>
>>>>> Tel: 401 863 1230
>>>>> Fax: 401 863 9182
>>>>> http://stat.brown.edu/~zwu
>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>
>>>
>
>
> --
> -------------------------------------------
> Zhijin (Jean) Wu
> Assistant Professor of Biostatistics
> Brown University, Box G-S121
> Providence, RI  02912
>
>
> Tel: 401 863 1230
> Fax: 401 863 9182
> http://stat.brown.edu/~zwu
>
>



More information about the Bioconductor mailing list