affylmGUI
Introduction
affylmGUI (Affymetrix linear modeling Graphical User Interface) is for Affymetrix data only.
If you have two-color cDNA data, you may wish to try limmaGUI.
affylmGUI Website Links
Main Webpage Contents
- Citation for affylmGUI
- Installing affylmGUI
- Download ONLY affylmGUI
- Windows Installation Notes
- Linux/Unix Installation Notes
- MacOS X Installation Notes
- Tcl/Tk Issues
- Running affylmGUI
affylmGUI is an implementation of a body of methodological research by the authors and coworkers. Please
cite the appropriate methodological papers whenever you use results from the limma software in a
publication. Such citations are the main means by which the authors receive professional credit for their
work.
Citing limma and affylmGUI in publications will usually involve citing one or more of the methodology
papers that the limma software is based on as well as citing the limma software package itself.
If you use limma/affylmGUI for differential expression analysis, please cite reference 1 which describes
the linear modeling approach implemented by lmFit and the empirical Bayes statistics implemented by eBayes,
topTable etc.
To cite the limma software itself please refer to reference 2 which describes the software package in the
context of the Bioconductor project and surveys the range of experimental designs for which the package can
be used, including spotspecific dye-effects. The pre-processing capabilities of the package are also
described but more briefly, with examples of background correction, spot quality weights and filtering with
control spots. This article is also the best current reference for the normexp background correction
method.
To cite the GC robust multiarray average (GCRMA) background correction method please refer to citation 3.
To cite the robust multiarray average (RMA) background correction method please refer to citation 4.
-
1. Smyth, G. K. (2004). Linear Models and Empirical Bayes Methods for
Assessing Differential Expression in Microarray Experiments. Statistical
Applications in Genetics and Molecular Biology Vol. 3 : Iss. 1, Article 3.
(Full text available here.)
-
2. Smyth, G. K. (2005). Limma: Linear Models for microarray data. In:
Bioinformatics and Computational Biology Solutions using R and Bioconductor,
R. Gentleman, V. Carey, S. Dudoit, R. Irizarry, W. Huber (eds.), Springer,
New York, pages 397-420. (Published 8 August 2005,
Publisher web site,
PDF)
-
3. Zhijin Wu1, Rafael A. Irizarry, Robert Gentleman, Francisco Martinez-Murillo, Forrest Spencer. (2004).
A Model Based Background Adjustment for Oligonucleotide Expression Arrays In the Journal of the American
Statistical Association. Volume 99, Pages 909–917.
-
4. Rafael A. Irizarry, Bridget Hobbs, Francois Collin, Yasmin D. Beazer-Barclay, Kristen J. Antonellis,
Uwe Scherf, Terence P. Speed. (2003). Exploration, normalization, and summaries of high density
oligonucleotide array probe level data In the Journal Biostatistics. Volume 4(2), Pages 249–264.
affylmGUI is distributed as open-source software under the GNU General Public License(GPL).
The best way to install affylmGUI is to use the biocLite function from the Bioconductor site.
However certain other software needs to be installed depending on the Operating System to be used.
See the notes for each particular OS.
The latest version of affylmGUI for R 2.5.0 available from
Bioconductor is 1.10.4
The latest version of affylmGUI for R 2.5.0 available from this site is 1.10.5.
The latest version of affylmGUI for R 2.6.0dev available from
Bioconductor Devel is 1.11.4
Files available from this site:
To view the latest changelog click here
To download affylmGUI and all its dependent packages, see the note for the appropriate Operating System below.
Download R from the R Project for Statistical Computing.
Follow the CRAN link and select a mirror near you.
Select the "Windows(95 and later)" link
Select the "base" link.
Select the R-2.x.y-win32.exe link. Save this file and run it on your computer.
It is suggested that you install it in the Directory C:\R (create it if necessary).
This usually doesn't need administrator privliges and should you ever get around
to compiling code, you will have less problems as some Windows compiling tools
don't like spaces in the directory names as there are in the "Program Files" directory name.
Download ActiveTcl for Windows from ActiveState.
Click the "Get ActiveTcl" button.
Click the "Download" button.
Click the "Continue" button.
Click "AS package" under the Windows heading.
Save the file, run it and install in C:\Tcl as prompted to do so.
Accept all the defaults and it will be installed in C:\Tcl and C;\Tcl\bin will be added to your PATH.
This install includes, among many others, the the Tcl/Tk packages, BWidget and Tktable,
which affylmGUI requires.
Install the required R packages.
You can install the affylmGUI package from this site or from Bioconductor.
See the statements under Download ONLY affylmGUI above to determine versions available.
Sometimes there maybe a later version of affylmGUI at this site, rather than at the Bioconductor site.
To install affylmGUI from this site, paste the following command into your R session:
>install.packages("affylmGUI",contriburl="http://bioinf.wehi.edu.au/affylmGUI")
However this doesn't install all the packages that maybe needed by affylmGUI.
The best way to install R packages is to use the biocLite function.
To install the correct version of a package for your version of R paste the following into your R console.
>source("http://www.bioconductor.org/biocLite.R")
Then run the biocLite function naming the package(s) you wish to install.
If you add the argument "dependencies=TRUE", then all the dependencies of the packages
will also be installed.
To install affylmGUI, with its dependencies, paste this command into the R console:
>biocLite("affylmGUI", dependencies=TRUE)
On a fresh install of R 2.5.1, this will install 48 packages!.
A minimal install of affylmGUI could be done using
>biocLite("affylmGUI")
This will install affylmGUI and its immediately dependent packages, limma, tcltk, affy,
Biobase and affyio.
Depending on how you use affylmGUI, you may also wish to install tkrplot, affyPLM, R2HTML and xtable.
To do this, use these commands:
>biocLite("tkrplot")
This installs just tkrplot.
>biocLite("affyPLM")
This installs affyPLM and dependencies, matchprobes, affydata and gcrma.
>biocLite("R2HTML")
This installs just R2HTML.
>biocLite("xtable")
This installs just xtable.
Download R from the R Project for Statistical Computing.
You will need to have administrator privliges to install this software in the default locations.
Follow the CRAN link and select a mirror near you.
Select the "Linux" link
Select the links to suit your flavour of Linux and your choice of Binary or Source code downloads.
Install the R software according to usual Linux procedures.
Linux usually has X11 and Tcl/Tk installed.
A common location for the installation of Tcl/Tk and X11 is /usr/lib.
Install Tktable from http://tktable.sourceforget.net.
Select File Distributions, In the "Latest File Releases" table, select the Download
button at the end of the tktable line. Click on the Tktable2.9.tar.gz file name and save it.
At the unix prompt where Tktable2.9.tar.gz has been saved, type the command:
gzip -d Tktable2.9.tar.gz
This will create Tktable2.9.tar. To untar this file, type the command:
tar -xvf Tktable2.9.tar
This creates the directory Tktable2.9
cd to the Tktable2.9 directory and read the README.txt file. Basically you need to
run the configure command. ie.:
./configure
or if you do not want the default locations, use the --prefix and --exec-prefix options.
A user without administrator access can Create a folder in their home area called "TkExtensions.
Use --prefix to set the location to this folder. an example configure command would be:
./configure --prefix=/home/users/keith/TkExtensions
Otherwise the default location is /usr/local.
Next install the program in the usual unix way, using these commands:
make
make test (OPTIONAL)
make install
Install BWidget from http://sourceforge.net/projects/tcllib/.
Click the Download button in the upper row of the Menu items.
In the "Latest File Releases" table, select the Download button at the end of the BWidget line.
Click on the BWidget-1.8.0.tar.gz file name and save it.
At the unix prompt where BWidget-1.8.0.tar.gz has been saved, type the command:
gzip -d BWidget-1.8.0.tar.gz
This will create BWidget-1.8.0.tar. To untar this file, type the command:
tar -xvf BWidget-1.8.0.tar
This creates the directory BWidget-1.8.0
cd to the BWidget-1.8.0 directory and read the README.txt file.
Basically you need to copy the directory to the same location that you installed the Tktable package.
Tcl/Tk uses the tcl variable "auto_path" to find Tktable and BWidget. Set this if Tcl/Tk is unable
to find either Tktable or BWidget. See below for details.
Now download affylmGUI and all its dependent packages.
See the instructions above under the heading Install the required R packages
to do this.
Download R from the R Project for Statistical Computing.
Follow the CRAN link and select a mirror near you.
Select the "MacOS X" link
Select the "R-2.5.0.dmg (latest version)" link. Read the instructions.
Save this file, double click and install R. This will install R and Tcl/Tk (compiled for X11). However it doesn't install BWidget and Tktable.
A useful site for R Mac users (actually more aimed at developers) is here.
affylmGUI uses Tcl/Tk compiled to use the X11 library for displaying its GUI.
There are three ways to install Tcl/Tk.
- Install with R as described above, but BWidget and Tktable packages are NOT installed.
- Install the complete ActiveState Tcl/Tk and get these Tcl/Tk packages and many more.
- Get the individual Tcl/Tk packages.
Downloading ActiveTcl for Mac OSX from ActiveState.
Administrator privliges would be required. It installs TkTable and BWidget, along with other Tcl packages.
These X11-compatible Tcl/Tk directories will generally be installed in /usr/local/lib/ or /usr/lib .
Be sure the location of tcl8.4 directory and the tk8.4 directory is in your PATH variable.
(don't use TclTkAquaBI because that gives you an Aqua version instead of an X11 version).
Note that the Aqua-compatible (but X11-incompatible) Tcl/Tk folders will generally be in
/Library/Tcl/. You don't want this location in your PATH before the above location.
Install X11 if it is not already installed.
If you need to install X11, get it from here
Install just the Tktable package.
Tktable is available in binary form here:
zipped-up binary version of Tktable2.9 built for MacOS X (PowerPC version)
zipped-up binary version of Tktable2.9 built for MacOS X (Intel version).
and here are the
Tktable 2.9 indiviual source files if you want them.
A user without administrator access can Create a folder in their home area called "TkExtensions.
Place the Tktable2.9 folder in this folder.
Install just the BWidget package.
BWidget is available form here:
A gz zipped version of BWidget-1.8.0 for all platforms
BWidget is not compiled code. It is all text files and can therefore run on any platform.
Place the BWidget-1.8.0 folder in the TkExtensions folder.
affylmGUI, on a Mac, adds the following locations to TclPath (The locations that it looks for Tcl applications)
"/sw/lib/tcl8.4", "/sw/lib", "/sw/lib/tk8.4" and "TkExtensions" in the HOME directory of the user.
affylmGUI, on a Mac and Linux, add the following locations to TclPath.
"/usr/local/lib","/usr/local/Tcl/lib","/usr/local/lib/Tcl","/usr/lib","/usr/lib/Tcl" and "/usr/local/ActiveTcl/lib".
So affylmGUI should be able to find these Tcl/Tk packages.
Install required R packages.
To download affylmGUI and all its dependent packages, see Install the required R packages
Testing Tcl/Tk in R
In an R session:
>library(tcltk)
This loads the tcltk interface. Remember to start the Xserver first if you are using a Mac.
Otherwise R freezes and you have to force quit and start R again.
Use the command
>ls("package:tcltk")
if you wish to see all the tcltk function names/commands (>230) available.
To test if Tcl/Tk is working, try creating a window called "tt". Use the command shown next:
>tt <- tktoplevel()
If a small window pops up somewhere on your screen (or wherever X11 is sending your graphical output),
then Tcl/Tk is installed and working on your operating system. If it doesn't appear then either Tcl/Tk
is not installed or not found or it may be an X11 issue.
On MS Windows X11 is not the issue as R uses the Microsoft Windowing software.
On Linux and Mac OSX you will need to have an X11 server running and you will needed to have defined
your DISPLAY environment variable. This can be done in a bash shell using a command like this:
%~>export DISPLAY=localhost:0
or if your are displaying on another computer, it maybe, for example:
%~>export DISPLAY=pc512.mydomain.edu.au:0
Check the value of your DISPLAY variable on your unix/Mac OSX computer using
%~>echo $DISPLAY
assuming you are using a bash shell.
To destroy this window either click its close icon, or use the R tcltk command
>tkdestroy(tt)
Note that "tt" was the name used to name the window in the tktoplevel command above.
To examine where R is looking for Tcl/Tk use the R tcltk commmand tclvalue to examine the value
of the auto_path variable:
>tclvalue("auto_path")
On my MS Windows box it reports:
[1] "{C:\\R\\R-2.5.0/Tcl/lib/tcl8.4} C:/R/R-2.5.0/Tcl/lib ./lib C:/R/R-2.5.0/Tcl/lib/tk8.4
C:/R/R-2.5.0/library/tcltk/exec"
Note that two backslashes are equivalent on one forward slash in the R world.
On my Linux box, it reports:
[1] "/usr/share/tcl8.4 /usr/share ./lib /usr/lib64 /usr/share/tk8.4
/usr/local/software/R-2.5.0/lib64/R/library/tcltk/exec"
Note that a package may modify auto_path. For example, load affylmGUI as shown below:
>library(affylmGUI)
If you do NOT choose to run affylmGUI from its dialogue start box, loading affylmGUI, on a Windows box,
will have added one more location to the auto_path.
Show the auto_path value again:
>tclvalue("auto_path")
[1] "{C:\\R\\R-2.5.0/Tcl/lib/tcl8.4} C:/R/R-2.5.0/Tcl/lib ./lib C:/R/R-2.5.0/Tcl/lib/tk8.4
C:/R/R-2.5.0/library/tcltk/exec C:/Tcl/lib"
If you actually run affylmGUI by giving the command:
>affylmGUI()
it adds more locations. Show the auto_path value again:
>tclvalue("auto_path") on Windows, it displays:
[1] "{C:\\R\\R-2.5.0/Tcl/lib/tcl8.4} C:/R/R-2.5.0/Tcl/lib ./lib C:/R/R-2.5.0/Tcl/lib/tk8.4
C:/R/R-2.5.0/library/tcltk/exec C:/Tcl/lib C:/Tcl/lib/tcllib1.9 C:/Tcl/lib/tklib0.4"
>tclvalue("auto_path") on Linux, it displays:
[1] "/usr/share/tcl8.4 /usr/share ./lib /usr/lib64 /usr/share/tk8.4
/usr/local/software/R-2.5.0/lib64/R/library/tcltk/exec /usr/local/lib /usr/local/software /usr/local/Tcl/lib
/usr/local/lib/Tcl /usr/lib /usr/lib/Tcl /usr/local/ActiveTcl/lib /home/users/lab0605/keith/TkExtensions"
You can add paths to auto_path using the R tcltk command "addTclPath". For example:
>addTclPath("C:/Users/keith/lib")
Examine the value of auto_path again:
> tclvalue("auto_path")
[1] "{C:\\R\\R-2.5.0/Tcl/lib/tcl8.4} C:/R/R-2.5.0/Tcl/lib ./lib C:/R/R-2.5.0/Tcl/lib/tk8.4
C:/R/R-2.5.0/library/tcltk/exec C:/Tcl/lib C:/Tcl/lib/tcllib1.9 C:/Tcl/lib/tklib0.4"
C:/Users/keith/lib"
Note that:
>?tcl
gives some brief help on tcltk commands available.
To begin, start up R (or RGui), and type
>library(affylmGUI)
Comments/Questions? Contact keith@wehi.edu.au.
Last modified:
275 June 2007
|
