[BioC] [R] Why SAM has totally diffent results in R2.1.1 and R2.4.0

Holger Schwender holger.schw at gmx.de
Tue Nov 21 12:45:01 CET 2006


Hi Wei Zhao,

I have also taken a look on the results of sam for R 2.1.1 and 2.4, and they are exactly the same, and there are actually no reason -- at least in the sam code -- why the results might/should differ. Did you set the random number generator, i.e. rand, to the same value? 

I, however, do not get the warning Martin gets. This was once a warning in the R check of siggenes which then suddenly disappeared after I have asked Seth what this warning means.

Best,
Holger


-------- Original-Nachricht --------
Datum: Mon, 20 Nov 2006 12:59:53 -0800
Von: Martin Morgan <mtmorgan at fhcrc.org>
An: "wzhao" <wzhao at mednet.ucla.edu>
Betreff: Re: [BioC] [R] Why SAM has totally diffent results in R2.1.1 and	R2.4.0

> Wei Zhao
> 
> This is not my software package and I do not use it regularly, but the
> following might help in getting help from others:
> 
> * Direct any follow-up to the bioconductor mailing list
> 
> * Make sure that you are using a coherent set of current software.
> 
> * Be more precise than 'totally different'.
> 
> * Clearly indicate the software you are using.
> 
> Following my own advice, I tried
> 
> > update.packages()
> 
> to update the installed packages; I might have done
> 
> > source("http://bioconductor.org/biocLite.R")
> > biocLite()
> 
> on a new installation. I then ran your suggested code (with
> library(siggenes) rather than multtest), with the following output:
> 
> > library(siggenes)
> Loading required package: multtest
> Loading required package: Biobase
> Loading required package: tools
> 
> Welcome to Bioconductor
> 
>     Vignettes contain introductory material. To view, type
>     'openVignette()' or start with 'help(Biobase)'. For details
>     on reading vignettes, see the openVignette help page.
> 
> Loading required package: survival
> Loading required package: splines
> 
> Changes in siggenes:
> Version 1.2.x: Added sam2excel, sam2html, and help.sam.
> Version 1.5.3: Now exprSet objects can also be used in the EBAM functions
>                for specifying 'data' and 'cl'.
> Version 1.7.1: siggenes can also handle ExpressionSet objects.
> 
> 
> Attaching package: 'siggenes'
> 
> 
>         The following object(s) are masked from package:multtest :
> 
>          .mt.BLIM 
> 
>         The following object(s) are masked from package:multtest :
> 
>          .mt.naNUM 
> > data(golub)
> Warning message:
> file 'golub.RData' has magic number 'RDA1'
>    Use of save versions prior to 2 is deprecated 
> > # golub.cl contains the class labels.
> > golub.cl
>  [1] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1
> 1 1 1
> > # Perform a SAM analysis for the two class unpaired case assuming
> > # unequal variances.
> > sam.out<-sam(golub,golub.cl,B=100,rand=123)
> > sam.out
> SAM Analysis for the Two-Class Unpaired Case Assuming Unequal Variances 
>  
>    Delta  p0   False Called      FDR
> 1    0.1 0.5 2424.77   2739  0.44276
> 2    0.7 0.5  262.21   1248  0.10508
> 3    1.3 0.5   12.11    507  0.01195
> 4    1.8 0.5    0.74    210  0.00176
> 5    2.4 0.5    0.01     76 6.58e-05
> 6    3.0 0.5       0     15        0
> 7    3.6 0.5       0      5        0
> 8    4.1 0.5       0      2        0
> 9    4.7 0.5       0      2        0
> 10   5.3 0.5       0      0        0
> 
> The software packages I'm using are:
> 
> > sessionInfo()
> R version 2.4.0 Patched (2006-10-04 r39581) 
> x86_64-unknown-linux-gnu 
> 
> locale:
> LC_CTYPE=en_US;LC_NUMERIC=C;LC_TIME=en_US;LC_COLLATE=en_US;LC_MONETARY=en_US;LC_MESSAGES=en_US;LC_PAPER=en_US;LC_NAME=C;LC_ADDRESS=C;LC_TELEPHONE=C;LC_MEASUREMENT=en_US;LC_IDENTIFICATION=C
> 
> attached base packages:
> [1] "splines"   "tools"     "methods"   "stats"     "graphics" 
> "grDevices"
> [7] "utils"     "datasets"  "base"     
> 
> other attached packages:
> siggenes multtest survival  Biobase 
>  "1.8.0" "1.11.2"   "2.29" "1.12.2" 
> 
> I found R 2.1.1 installed on a Windows laptop I have. The results of
> sam.out are nearly identical (the formatting in the R 2.1.1 version
> did not use scientific notation) to those presented
> above. Unfortunately I do not have access to the laptop at this
> instant, so cannot provide the detail above. The installed software
> was, I think, siggenes 1.2.17, so not exactly comparable to
> you. 
> 
> There are some suspicious items in the above, particularly the warning
> about 'Use of save versions prior to 2 is deprecated' and perhaps also
> the objects being 'masked'. It is hard to know whether these are worth
> pursuing as a source of your problems. How does your output differ?
> 
> siggenes seems to be actively and conscientiously maintained, so a good
> source for help, if the mailing list fails you, might be the package
> author; try
> 
> > library(help=siggenes)
> 
> for contact information.
> 
> Hope that helps,
> 
> Martin
> 
> "wzhao" <wzhao at mednet.ucla.edu> writes:
> 
> > Hi,
> >
> >    I am using SAM (from siggenes_1.2.11 package) method to select genes
> from
> > a microarray data set. After installing the latest R2.4.0 on my
> computer, to
> > my surprise the results are totally different from that calculated using
> > R2.1.1. Even the example code doesn't work the same way under these two
> > versions of R. Does anybody know what is going on? Thanks for any
> > suggestions.
> >
> >  
> >
> >  
> >
> >  library(multtest)
> >  data(golub)
> > # golub.cl contains the class labels.
> >   golub.cl
> > # Perform a SAM analysis for the two class unpaired case assuming
> > # unequal variances.
> >   sam.out<-sam(golub,golub.cl,B=100,rand=123)
> >   sam.out
> > # Obtain the Delta plots for the default set of Deltas
> >   plot(sam.out)
> >
> >  
> >
> >  
> >
> > Best
> >
> > Wei Zhao
> > Postdoctoral Research Fellow
> > Department of Human Genetics and Biostatistics
> > University of California, Los Angeles
> > Gonda Neuroscience and Genetics Research Center 
> > 695 Charles E. Young Drive South, Box 708822 
> > Los Angeles, CA 90095-7088, USA
> >
> > Email: wzhao at mednet.ucla.edu
> > Tel:(310)8251677
> >
> >
> > ----------------------------------------------------------
> > IMPORTANT WARNING:  This email (and any attachments) is only intended
> for the use of the person or entity to which it is addressed, and may contain
> information that is privileged and confidential.  You, the recipient, are
> obligated to maintain it in a safe, secure and confidential manner. 
> Unauthorized redisclosure or failure to maintain confidentiality may subject you
> to federal and state penalties. If you are not the recipient, please
> immediately notify us by return email, and delete this message from your
> computer.
> > ----------------------------------------------------------
> >
> > 	[[alternative HTML version deleted]]
> >
> > ______________________________________________
> > R-help at stat.math.ethz.ch mailing list
> > https://stat.ethz.ch/mailman/listinfo/r-help
> > PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> > and provide commented, minimal, self-contained, reproducible code.
> 
> -- 
> Martin T. Morgan
> Bioconductor / Computational Biology
> http://bioconductor.org
> 
> _______________________________________________
> 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