[BioC] FW: Problem in masking individual probes to calculate expression values

Jenny Drnevich drnevich at illinois.edu
Tue Feb 9 04:43:18 CET 2010


Hi Farhat,

I've never actually used it to remove single probes, so I'm not sure 
I can help you all that much. I'm not the original author of the 
RemoveProbes() function, so I'm not 100% certain how everything 
works. A few suggestions for you to try:

1) Exactly how many probes are you talking about when you say "it is 
working well to mask few probes"? Did you only try doing probes from 
one probe set? If you were trying probes from several probe sets at a 
time and that worked, but the entire set of 25,000 probes didn't 
work, then you may have some incorrect probe names that are causing 
the error. Try cutting the list in half and seeing if one half works 
but the other half doesn't. If there are a few incorrect probe names, 
it may take some sequential halving of the list to find them.

2) I don't think you can call RemoveProbes() more than once without 
resetting the cdf environment. Did you just try RemoveProbes() with a 
few probes, then immediately try RemoveProbes() with the larger set? 
That may be the problem. Remember this when you're checking #1 above 
- you have to reset the cdf back to it's original form before trying 
another set of probes.

3) If #1 and #2 don't get you anywhere, you'll need to step through 
the code of RemoveProbes() to see where the error pops up. Do

 > debug(RemoveProbes)

then run RemoveProbes() as you normally would. That should spit out 
the code and open the debugger. Basically just hit enter to evaluate 
one line of code at a time. When you get the error, then you know 
what line caused it. This would help me track down the problem.

4) One quick check - when you put all 25000 probes in listOutProbes, 
you're putting them in as a single character vector, and not some 
other sort of object, correct?

HTH,
Jenny

P.S. Keep everything on the mailing list for the record.

At 04:12 PM 2/6/2010, Memon, Farhat N wrote:
>Hi Jenny,
>
>Thanks for your reply. I was ill so couldn't thank you earlier. I am 
>using the format 117_at1 and so on. it is working well to mask few 
>probes but ..............
>
>I need to mask about 25000 probes before applying GCRMA using your 
>code 
>(https://stat.ethz.ch/pipermail/bioconductor/2006-September/014242.html). 
>But when I define all the 25000 probes in the listOutProbes and call 
>RemoveProbes(), it always gives the following error :
>
>   Error in ans[[i]][, i.probes] : incorrect number of dimensions
>   Error in unlist(indexProbes(newAB, "pm")) :
>     error in evaluating the argument 'x' in selecting a method for 
> function 'unlist'
>
>
>Is there any way to mask such a huge number of probes using your code?
>
>Thanks,
>Farhat Memon
>________________________________________
>From: Jenny Drnevich [drnevich at illinois.edu]
>Sent: 26 January 2010 15:43
>To: Memon, Farhat N; bioconductor at stat.math.ethz.ch
>Subject: Re: [BioC] FW: Problem in masking individual probes 
>to  calculate expression values
>
>Hi Memon,
>
>RemoveProbes() requires the probe name, in the format of 117_at1 or
>117_at9, not the x,y coordinates. I'm not sure there's an easy way to
>get from x,y coordinates to probe names, but some combination of
>indexProbes() and indices2xy() should work. You'll need an AffyBatch
>object of the raw data to test it, called "rawdata" in my code below.
>
>  > x <- unlist(indexProbes(rawdata,genenames="117_at"))
>  > x     # this will give you the indices of the pm probes for probe
>set 117_at along with their names
>
>  > y <- indices2xy(x, abatch=rawdata)
>  > y     # this will give you the x,y coordinates along with the
>correct probe names
>
>HTH,
>Jenny
>
>
>
>
>
>At 08:56 AM 1/18/2010, Memon, Farhat N wrote:
>
> >Reference:
> >https://stat.ethz.ch/pipermail/bioconductor/2006-September/014242.html
> >
> >Hi,
> >
> >To calculate gcrma values of HG_U133_Plus_2 arrays, I am using Jenny
> >Drnevich's code (see the link above) (Thanks to Jim who provided
> >that link to me). This code is to mask either probesets or
> >individual probes while calculating gcrma values. I have
> >successfully run the code to mask some of the probe sets. Now, I
> >want to mask individual probes.
> >
> >For example, I want to calculate gcrma value of probe set 117_at
> >without a probe 60:815 and the probe set 121_at excluding three of
> >its probes 801:783, 228:969, 268:273 (the numbers are the
> >probes'  x:y positions).
> >
> >According to the code (in the above link), to call function
> >RemoveProbes(), I need to provide list of probe to be masked but it
> >is not mentioned in which format one should provide probe list.
> >
> >I used the following code that is not working:
> >
> >probe_list=c("117_at:60:815","121_at:801:783", "121_at:228:969",
> >"121_at:268:273")
> >
> >cleancdf=cleancdfname("HG_U133_Plus_2", addcdf=FALSE)
> >ResetEnvir(cleancdf)
> >RemoveProbes(listOutProbes=probe_list, listOutProbeSets=NULL, cleancdf)
> >
> >I am getting error while masking probe. If anybody used this code to
> >mask individual probes, please let me know the format of the probe list.
> >
> >Thanks for your help in advance.
> >
> >Regards,
> >Farhat
> >
> >_______________________________________________
> >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
>
>Jenny Drnevich, Ph.D.
>
>Functional Genomics Bioinformatics Specialist
>W.M. Keck Center for Comparative and Functional Genomics
>Roy J. Carver Biotechnology Center
>University of Illinois, Urbana-Champaign
>
>330 ERML
>1201 W. Gregory Dr.
>Urbana, IL 61801
>USA
>
>ph: 217-244-7355
>fax: 217-265-5066
>e-mail: drnevich at illinois.edu

Jenny Drnevich, Ph.D.

Functional Genomics Bioinformatics Specialist
W.M. Keck Center for Comparative and Functional Genomics
Roy J. Carver Biotechnology Center
University of Illinois

330 ERML
1201 W. Gregory Dr.
Urbana, IL 61801

Ph: 217-244-7355
FAX: 217-265-5066
Email: drnevich at uiuc.edu



More information about the Bioconductor mailing list