[BioC] probe masking

[James MacDonald]

Hi Donna,

Donna wrote:
>  
> James MacDonald <jmacdon at ...> writes:
> 
>> Hi Donna,
>>
>> See the function posted by Jenny Drnevich about half way down this post:
>>
>>
> http://article.gmane.org/gmane.science.biology.informatics.conductor/9887/match=probe+mask
>> Best,
>>
>> Jim
>>
>> Donna Toleno wrote:
>>> Hello,
>>>
>>> I am looking for an answer which I believe has been discussed on the list
> previously. I would like to know how
>> to modify the cdf environment in order to mask out certain probes for a
> between species analysis.
> 
> 
> Thank you for your reply Jim. It has been very helpful. I was able to remove the
> unwanted "masked" intensity data by changing the appropriate intensity values to
> NA. Then I found that this is not a solution to my problem because RMA will not
> accept NA values. So I'm back to modifying the cdf. 
> 
> I used the code you pointed me to as well as some related code posted by Ariel
> Chernomoretz. When I try things line by line, I have found that there is still a
> problem with the following line:
> 
> 
>> a<-get(pset[i],env=get(cdfpackagename))
> Error in get(x, envir, mode, inherits) : invalid 'envir' argument
> 
> 
>> cdfpackagename
> [1] "mouse4302cdf"
>> pset[i]
> [1] "1448591_at"
> 
> I wonder what the following lines are for?
> 
>      #a<-get(pset[i], env=get(cdfpackagename))

This line gets the pm and mm indices from the cdf package:

 > a<-get(pset, env=get(cdfpackagename))
 > a
           pm     mm
  [1,] 174302 175304
  [2,] 230237 231239
  [3,] 530308 531310
  [4,]  37779  38781
  [5,] 763355 764357
  [6,] 230427 231429
  [7,]   7137   8139
  [8,]  29700  30702
  [9,] 987507 988509
[10,] 564747 565749
[11,] 250297 251299

>      #a<-a[-iout,]

This line removes the rows you don't want for that probeset

>      #assign(pset[i],a,env=get(cdfpackagename))

This line puts what remains for that probeset back into the cdf environment.

> 
> I thought that I could at least remove the whole probesets without the above
> problematic lines.
> 
> so when I try
> 
>> rm(list=listOutProbeSets,envir=get(cdfpackagename))
> 
> I get a similar error.
> 
> Error in remove(list, envir, inherits) : invalid 'envir' argument
> 
> So I guess my question is what is the valid environment argument?

For me, that envir argument works. How about trying just these lines of 
code on a newly started R session:

cdfpackagename <- "mouse4302cdf"
pset <- "1448591_at"
library(mouse4302cdf)
a<-get(pset, env=get(cdfpackagename))
a

If that works, then maybe your R session got buggered somehow, and you 
can then try again.

Just to make sure; you are trying to remove probes, not probesets? If 
you want to remove complete probesets you should look at the altcdfenvs 
package instead.

Best,

Jim


> 
>> listOutProbeSets
> [1] "1415897_a_at" "1434465_x_at" "1439066_at"  
> 
> Could the problem be due to syntax differences. I am using R and Bioconductor on
> Windows?
> I made a few syntax changes already including a change from "xy2i" to
> "xy2indices" based on an error message. Unfortunately, the help files for rm and
> for get were not very helpful.
> 
> Any comments would be appreciated.
> 
> Thanks,
> 
> Donna
> 
> _______________________________________________
> 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

-- 
James W. MacDonald, MS
Biostatistician
UMCCC cDNA and Affymetrix Core
University of Michigan
1500 E Medical Center Drive
7410 CCGC
Ann Arbor MI 48109
734-647-5623