Appendix D. Use of MAExplorer as a stand-alone application

The MAExplorer program is used primarily as a stand-alone Java Application. The same Java program may also be used as a Java Applet that runs within a Web browser. Both the applet version and the stand-alone version may be run on a stand-alone computer without an Internet connection if the microarray database data files reside on that machine. The stand-alone Java version is more robust and has some important advantages discussed in the Introduction and
Appendix E.2.

This section discusses the installation of MAExplorer as a stand-alone application on a variety of computers. Since Java is portable between Microsoft Windows (95/98/NT/2000/XP), Macintoshes, Linux, Solaris, etc., it is possible to freely download and install MAExplorer and Cvt2Mae on your computer and run it as an application program.

There is a discussion on using it with other arrays (Appendix C) that requires editing data files for use with MAExplorer. An array data conversion tool is being constructed which will automate this process in the future.

  1. Stand-alone applications can utilize huge amounts of memory (for example, we were able to use >80Mb on a 128Mb Windows NT system). This means that an analysis of large numbers of hybridizations and complex statistical computations could be done locally.

  2. The stand-alone version can do local disk I/O. This means that data-mining sessions may be saved and restored locally to continue a data-mining session at a latter time. Plots may be saved as GIF (.gif) files, and text windows as ASCII text (.txt) files. Hybridized sample quantification files slowly downloaded from a microarray database Web server may be cached on the local machine - saving time when the data session is resumed at a later time. Users could then add GIPO conforming hybridized sample quantification files on their desktop and perform a data-mining analysis of hybridized data from multiple sources - mixing data from public servers as well as private data. This latter option would be most useful using commercial arrays.

  3. Users may download a self-extracting Java application. This is available on the MAExplorer Web site for a variety of computer systems and includes a Java Virtual Machine (JVM) for your operating system. [The JVM is only available on the installers on the LECB/NCI web site, it is not available on the SourceForge Web site because of space considerations. However, if by some chance your computer does not have a JVM, this should not be a problem since the JVMs (called JDK's) are available at various Web sites.] This JVM does not interfere with any other JVMs you may already have installed. However, it guarantees optimal operation of MAExplorer. This makes the stand-alone application more robust than the Applet as well as being simple to install.

  4. Having more memory implies allows taking advantage of more sophisticated statistical, clustering and data-mining methods, as well as working with large numbers of samples with larger numbers of spots. Some Web browsers will not give you access to large amounts of memory - even if it is available on your computer.

  5. Having the stand-alone version on the user's computer allows them to configure a local database for their own data.

D.1 Installing MAExplorer as stand-alone application

It is possible to download MAExplorer for data-mining either stand-alone on an your personal computer or in setting up a Web site for publishing your array data (like the
MGAP site).

Figure D.1 Web page showing options for installing MAExplorer as a stand-alone application. Installers are available for Windows95/98/NT/2000/XP, Mac OS, Solaris, Linux, Unix, and other Java enabled platforms.

D.2 Downloading MAExplorer for stand-alone use with other arrays

After MAExplorer is installed, you may start it by clicking on a startup icon for one of the saved databases (Windows) or running the program (Unix, Macintosh). In Windows, you may also start it from the Window's "Start" menu. This stand-alone version enables you to save data on your local machine and to run MAExplorer independent of an Internet connection as well as other features that are discussed throughout this manual.

You first need to download MAExplorer for your particular type of operating system. These include Windows 95/98/NT/2000/XP, MacOS for Power PC, Sun Solaris, HP-UX, other Unix versions (e.g. Linux, etc). The Windows, MacOS and Solaris versions include a Java run-time (Java Virtual Machine) that works with MAExplorer. We recommend you download the full distribution for your computer (which includes a recent Java virtual machine (JVM) if it exists). This insures proper operation of MAExplorer and does not interfere with other Java applications you might have installed or will install.

This installation process uses a commercial "Java Installer" (InstallAnywhere(TM) by ZeroG Inc.) that requests you "Grant" it permission to save the installation on your computer. It will suggest where to install it or you can install it wherever you want. For example, in Windows it may suggest saving it in C:\Program Files\MAExplorer\ - you can specify an alternative directory if the default disk does not have much free space left.

  1. Go to download MAExplorer to begin the installation.
  2. With Netscape, Click on "Grant" to questions "Starting programs stored on your computer" and "Reading, modification, or deletion of nay of your files". With Internet Explorer, click on "Yes" in response to the popup window "Do you want to install and run ...". [All data files are kept in the same directory tree where the MAExplorer program is kept. MAExplorer does not use any other directories on your computer and does not affect other Java programs on your computer.]
  3. Either click on the default "Download ..." button that was automatically determined for your machine or scroll down and click on a "download" hypertext link for the operating system you want to download. If it is the latter, it pops up a "Save as" window for you to select a temporary directory in which to put the "installMae.exe" (Windows), "installMae.bin" (MacOS and Unix systems), etc. In Windows this could be "C:\Temp" or in Unix "/var/tmp", etc. After you have installed MAExplorer, you can delete this file.
  4. After the file was downloaded on your machine, go to that directory and start that file (e.g. click on it on a Windows or MacOS system). [If you wish, you may exit your Web browser at this point since it is no longer needed].
  5. When the installer starts, it pops up a window "Preparing to Install" and gives you the option of selection a language. Click on the "OK" button.
  6. It then pops up an "Introduction" window. Click on "Next".
  7. It then pops up a "Choose Install Folder". You have the option of specifying a different directory. E.G. (on Windows) if it says: C:\Program Files\MAExplorer and you don't have space on disk C: but do on disk E: and want to put it into its own directory, you could specify it as E:\MAExplorer. Then click on the "Install" button.
  8. It then pops up a new window with "Install Complete". Click on the "Done" button. You are now ready to run MAExplorer.

D.2.2.1 Subsequent updating of only the MAExplorer JAR file from the MAExplorer server

Generally, you can greatly shorten the subsequent download of new versions of MAExplorer. This will work most of the time as most often this is the only change in the distribution (not including the documentation). If this causes problems, it may mean that the other files have changed and you may have to reinstall the complete distribution as in Appendix D.2 above.
  1. Download just the MAExplorer.jar file.
  2. Use this to replace the previous version of the MAExplorer.jar file on your system with this file.

Additional instructions for Installing MAExplorer on Sun Solaris

These instructions are valid for use with both TWM and CDE window managers. If you have problems with the Sun installer, you may need to update your Solaris OS system patches using a recent patch set.

It may involve more than a single patch. It is the latest Recommended Patch Cluster from Sun. We   STRONGLY recommend having your System Admininistrator do this for you if you have not done this before. Point your Web browser to:
and choose the appropriate patch set for the version of Solaris (2.6, 7, or 8, etc.) that you are running. Do not choose any of the x86 versions unless you are running Solaris x86. Click on either the Download HTTP option or Download FTP option, and click the GO button to download the patch set.

  1. Download Solaris version of MAExplorer by clicking on the correct link.
  2. After downloading, open a shell or xterm window then, cd to the directory where you downloaded the installer.
  3. At the prompt type: "sh ./installMAExplorer.bin". this will run Install Anywhere.
  4. The InstallAnywhere program should now be running, follow the instructions to install MAExplorer.
  5. After MAExplorer is installed click on the Exit key to end the Install Anywhere program. Then cd to the directory where it was installed and type "MAExplorer&" to run MAExplorer.
  6. To uninstall MAExplorer: cd into /UninstallerData subdirectory and at the Unix prompt type Uninstall_MAExplorer. This will bring up a window to prompt you to uninstall MAExplorer. Click on uninstall button.
If you requested it, note that a Java virtual machine is included with this download. It will run automatically when you run the shell script.

D.2.2 Enabling MAExplorer to fetch data from a microarray Web server

When run as a stand-alone application, MAExplorer can be set up to read data from a MAExplorer enabled Web database server and to cache the data to your local computer. The following lists the three variables that should be set in the .mae startup database file (see full table in discussion on setting up the configuration file). These variables may be set from the Edit menu Preferences submenu Use Web DB and Web DB data caching toggle switches, and the URL of the database by the Set Web DB command.

Table D.2.2 Parameters specifying Web database access. These optional parameters may be used for control access to a microarray Web database. The example values are shown for the MGAP database.

Parametery Value DataType Comments
(opt) enableFIOcaching TRUE boolean enable caching data files from Web server on local compute
(opt) saCodeBase String Web database to use to get the data
(opt) useWebDB TRUE boolean set get data from a Web database

D.2.3 Enabling MAExplorer cache to save Web data on local computer

When run as a stand-alone application, MAExplorer can be set up to cache data from a MAExplorer enabled Web database server to your local computer. You can enable caching for future access using the Edit menu Preferences submenu Web DB data caching toggle switch.

D.3 Starting MAExplorer by clicking on a .mae file

Once the MAExplorer stand-alone application has been installed and registered with the operating system, then for Windows (and other operating systems when you can register files for startup) just click on the icon for any file with a .mae file extension to start MAExplorer on that data. For Unix systems you can start MAExplorer on a .mae file by:
   (installation path)/MAExplorer MAE/(some startup file).mae

You can let your Unix system find MAExplorer by putting it in your path variable in your login or shell startup script.

   set path = ($path <MAExplorer installation path>)

Then, you would start it by specifying the startup file residing in the MAE/ subdirectory as:

   MAExplorer MAE/(some startup file).mae

There is a set of sample MGAP .mae files in the MAE subdirectory in the downloaded installation.

D.4 The data file format for .mae startup files

When MAExplorer is used as a stand-alone application, it first reads a (tab-delimited) startup file. This file contain the names of the hybridized samples to be loaded as well as some of the additional parameters listed here. Note most of these parameters may be specified in the configuration file as defaults and therefore do not need to be included in the .mae file unless you wish to override the configuration values. (See Table C.5 tables for a list of these parameters).

The .mae startup files are simply tab-delimited ASCII files with a .mae file extension. These could be created or edited either manually (e.g. using Microsoft Excel and saving the file as a tab-delimited file) or by various database programs (eg. the NCI/CIT mAdb program, the MAExplorer Cvt2Mae program being developed). They may also be generated by MAExplorer (File:Database:Save as file DB). The .mae file form consists of two tab-delimited columns containing fields Name and Value. These field names appear in the first row. This is followed by instances of the various parameters. A simple .mae file is shown in the following table Table D.4 using a 4 sample Lactation database subset from the MGAP database. Although any of the configuration file values can be specified in the .mae file, we list some of the more common optional parameters are indicated in Table D.4.1.

Table D.4 The Minimum data required entries for .mae startup files. These entries are shown with example values from some of the data in the MGAP demonstration database. Each sample is specified by an imagei name.

Name Value
image1 C57B6-L1-30min
image2 C57B6-L3-1hr
image3 C57B6-L10-29hrs-1
image4 Stat5a.--.L1-30min

Table D.4.1 Some of the common optional entries for .mae startup files. These entries are shown with example values for the 4 samples in Table D.4. See (Appendix C.5 for lists of many other options.

Name Value DataType Comments
(opt) maxPreloadImages 4 int overide the number of samples (called images) to actually load. This may be less than the number of image entries.
(opt) configFile MaExplorerConfig-MGAP.txt String name of Configuration file if not MaExplorerConfig.txt
(opt) dataBase MGAP DB String name of this specific database
(opt) dbSubset Pregnancy 13 days: C57BL/6 vs. stat5a (-,-), 8 samples String title for this subset name of the database
(opt) Xlist 1,2,3 String hybridized samples for initial HP-X 'set'. Corresponding to image1, image2, etc. Empty if not defined - may be defined using the Choose HP-X(Y,E) in the File menu.
(opt) Ylist 4 String hybridized samples for initial HP-Y 'set'. Corresponding to image1, image2, etc. Empty if not defined - may be defined using the Choose HP-X(Y,E) in the File menu.
(opt) Elist 1,2,3,4 String hybridized samples for initial HP-E 'list'. Corresponding to image1, image2, etc. Empty if not defined - may be defined using the Choose HP-X(Y,E) in the File menu.
(opt) classNameX C57B6 lactation (days 1,3,10) String Experimental class name for the HP-X 'set' of hybridized samples
(opt) classNameY Stat5a (-,-) lactation day 1 String Experimental class name for the HP-Y 'set' of hybridized samples
(opt) noMsgReporting TRUE boolean If set TRUE, used with Applet only to not send loading status message.
(opt) reuseXYcoords FALSE boolean If set TRUE and the quantified data files have the (x,y) coordinates for each spot, then use the same coordinates for all subsequent data files so that the arrays can be superimposed (for Flickering two HPs).
(opt) usePseudoXYcoords FALSE boolean If set TRUE, force MAExplorer to generate pseudoarray (X,Y) spot coordinates and ignore (X,Y) data in the quantified spot files if it exists. This will be set to TRUE automatically if there are no (X,Y) data fields in the quantified spot files.

D.5 Using MAExplorer as an Applet on your computer

It is possible to create a Web site to publish users data using the MAExplorer.jar file to support your private microarray database Web site. (Note that you can also get the MAExplorer.jar file from the directory where you installed MAExplorer on your computer). You might choose to mimic the way we did the MGAP Web site or organize it differently. You need to do the following:
  1. Create a a sub-directory dir/ in your htdocs/ Web server tree (where dir/is what you called the directory).
  2. Create dir/Config and dir/Quant sub-directories.
  3. Copy the MAExplorer.jar file into dir/.
  4. Edit configuration files and copy them to dir/Config sub-directory.
  5. Copy quantification files into the dir/Quant sub-directory.
  6. Add HTML Web pages containing <APPLET> code (see example below).

The following is a simple example of HTML code containing an applet which will invoke MAExplorer. You may add other options with PARAMs in the Applet (or for that matter in the the .mae startup file) that overide any options normally specifed in the Configuration file (See Appendix C.5).

<TITLE>MAExplorer Startup: C57B6 Pregnancy vs Lactation</TITLE>

<H2>MAExplorer Startup: C57B6 Pregnancy vs Lactation</H2>

This startup database will start the MAExplorer. It contains a subset
of the database consisting of four C57B6 mammary development
hybridized samples (HP): two each for pregnancy and lactation.

<APPLET CODE=MAExplorer.class  ARCHIVE=MAExplorer.jar
        WIDTH=10 HEIGHT=10 ALIGN=absmiddle>
  <PARAM NAME=configFile VALUE=MaExplorerConfig-MGAP.txt>
  <PARAM NAME=dbSubset VALUE="C57B6 pregnancy vs lactation">
  <PARAM NAME=image1 VALUE=C57B6-p13.1>
  <PARAM NAME=image2 VALUE=C57B6-L1-30min>
  <PARAM NAME=image3 VALUE=C57B6-p13.2poly-A>
  <PARAM NAME=image4 VALUE=C57B6-L1-total>
  <PARAM NAME=Xlist VALUE=1,3>
  <PARAM NAME=Ylist VALUE=2,4>
  <PARAM NAME=Elist VALUE=1,3,2,4>
  <PARAM NAME=classNameX VALUE=Pregnancy>
  <PARAM NAME=classNameY VALUE=Lactation>
(Sorry, you need a Java-capable browser to view this.)

D.6 List of startup .mae files included in the download installation

This is a list of the .mae startup files from the
MAExplorer MGAP database. These are included with the distribution for use in the tutorials, etc. They include data from 50 hybridized samples of mouse mammary breast tissue including normal and some knockout samples. There are about 1700 duplicate clones on the arrays which are membranes printed by Research Genetics and hybridized by the MGAP group. See the MGAP home page for more information on these samples. The .mae files are available as separate files The data is also packaged as a zip file, and a Unix tar file
.mae file name X vs Y comparison # of hybridizations
C57vsDevModels-15probes.mae HP-XY is C57B6 vs developmental models 15
C57vsDevModels-15probes-cache.mae HP-XY is C57B6 vs developmental models (with cache) 15
C57vsDevModels-38probes.mae HP-XY is C57B6 vs developmental models. HP-E is all samples 38
Lact1-C57vsStat5a-38probes.mae HP-XY is C57B6 Lactation day 1 vs Stat5a (-,-). HP-E is all samples 38
Lact1vs10-10probes.mae HP-XY is C57B6 Lactation day 1 vs Lactation day 10. HP-E is all samples 10
Lact1vs10-38probes.mae HP-XY is C57B6 Lactation day 1 vs Lactation day 10. HP-E is all samples 38
Lact-C57vsStat5a-5probes.mae HP-XY is C57B6 Lactation day 1 vs Stat5a (-,-) 5
Lact-C57vsStat5aCEBPnull-19probes.mae HP-XY is C57B6 Lactation day 1 vs Stat5a (-,-) and CEBP-null, HP-E has samples of other tissues 19
MAEstartupDefault.mae none none
Preg13day-C57vsStat5a-19probes-cache.mae HP-XY is C57B6 Pregnancy day 13 vs Stat5a (-,-). HP-E has samples of other tissues (with cache) 19
Preg13day-C57vsStat5a-19probes.mae HP-XY is C57B6 Pregnancy day 13 vs Stat5a (-,-). HP-E has samples of other tissues 19
Preg13day-C57vsStat5a-38probes.mae HP-XY is C57B6 Pregnancy day 13 vs Stat5a (-,-). HP-E is all samples 38
Preg-C57vsStat5a-4probes.mae HP-XY is C57B6 Pregnancy day 13 vs Stat5a (-,-) 4
Preg13VsLact1-18probes.mae HP-XY is C57B6 Pregnancy day 13 vs Lactation day 1. HP-E is all samples 18
Preg13VsLact1-38probes.mae HP-XY is C57B6 Pregnancy day 13 vs Lactation day 1. HP-E is all samples 38
Preg-C57vsStat5a-8probes.mae HP-XY is C57B6 Pregnancy day 13 vs Stat5a (-,-) 8
Preg13day-Stat5aVsCEBP-null-38probes.mae HP-XY is C57B6 Lactation day 1 vs Stat5a (-,-) and CEBP-null, HP-E is all samples 38
reuseXY-Preg13day-C57vsStat5a-38probes.mae HP-XY is C57B6 Pregnancy day 13 vs Stat5a (-,-). Use XY coords of first probe for remainder for flickering. HP-E is all samples 38
reuseXY-Preg-C57vsStat5a-8probes.mae HP-XY is C57B6 Pregnancy day 13 vs Stat5a (-,-). Use XY coords of first sample for remainder for flickering 8
MGAP-50samples.mae C57B6 day 13 preg. vs day 1 lact., 50 samples 50
OCL-P13L1L10Stat5a--15probes.mae replicates of C57B6 (pregnancy day 13, lactation days 1 and 10, and stat5a(-,-) 15 samples. The database also includes 4 additional condition sets of this data and an Ordered Condition List of the 4 conditions (in the State/ directory). This may be used to demo the OCL F-test filter. 15