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

[Martin Morgan]

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