Table of Contents

1. Introduction
   1. Algorithm
   2. Features
2. Download and Installation GifiArray Analyzer is distributed in 2 forms. Both of them work independently.
   1. GifiArray Analyzer with network database access
   This version of GifiArray Analyzer is easy to install. It connects directly to the server in our lab, therefore the running speed is affected by the network bandwith and the number of users on the server.
   2. GifiArray Analyzer with local database access
   This version of GifiArray Analyzer requires some experience with MySQL server to set up correctly.
3. Running GifiArray Analyzer
1. Input Data Format
2. A Walk-through with Sceen Shots
1. Basic Data Analysis
2. Gene Ontology/MIPS Analysis
3. Threashold Setting
4. Other Features i.e. print, export ..
4. How to cite?
5. Contact Information

1. Introduction

Gifi Array Analyzer is a Java application for integrated analysis of DNA microarray data and gene functional classification data.

1.1 Algorithm

Gifi Array Analyzer implements Homals (Homogeneity Analysis by means of Alternating Least Square). Homogeneity analysis is a graphical, multivariate statistical method. It is the basic component of the Gifi System.

1.2 Features

Features include: interactive data visualization, integrative analysis of gene expression profiles and functional categories, Gene Ontology Analysis, automatically downloading, parsing database source files of UniGene, LocusLink, SGD and Gene Ontology, and creating and updating a local relational database containing information extracted from the database source files.

2. Download and Installation

2.1 GifiArray Analyzer with network database access

2.1.1 System Requirements:

Gifi Array Analyzer is written in Java and has been tested on Windows 2000/XP and Redhat Linux 8.0. GifiArray Analyzer requires the following software to run successfully.

• Internet access
• Java Runtime Environment (JVM) 1.4.1 or later version, which can be downloaded from http://java.sun.com/products/archive/index.html
  To find out if you have JVM on your computer and its version, run 'java --version' from a command prompt. If it shows 'command unfound/unrecognized', you will need to download it from the above URL and install it.

2.1.2 Download GifiArray Analyzer:

• Gifi Array Analyzer for Windows 2000/XP can be downloaded from here
• Gifi Array Analyzer for Redhat Linux can be downloaded from here

2.1.3 Installation:

• Download the distribution file to a directory of your choice and unzip it. You will find a subdirectory named 'GifiArray_Analyzer' created in your installation directory.

2.2 GifiArray Analyzer with local database access

2.2.1 System Requirements

Gifi Array Analyzer is written in Java and has been tested on Windows 2000/XP and Redhat Linux 8.0.

• 1 GB of disk space are required to accomodate the database and input data files.
• GifiArray Analyzer requires the following software to run successfully.
  • Java Runtime Environment (JVM) 1.4.1 or later version, which can be downloaded from http://java.sun.com/products/archive/index.html
    To find out if you have JVM on your computer and its version, run 'java --version' from a command prompt. If it shows 'command unfound/unrecognized', you will need to download it from the above URL and install it.
  • MySQL Server and Client 3.23.56 or later version, which can be downloaded from http://www.mysql.com/downloads/mysql-4.0.html.
    You can follow these steps below if you need a new installation:
    • For Windows:
      

      Unzip the distribution file to a temporary directory;
      Run 'setup.exe' and install it in the default directory 'c:\mysql';
      Open a command prompt,
      > cd \mysql\bin
      > mysqld --install
      > net start mysql
      

    • For Redhat Linux:
      

      MySQL server is usually installed by default if you chose a 'server level' installation when your OS is initially installed.
      You can check if that is the fact by issuing command 'rpm -q mysqld'.
      

    After a new installation of MySQL server, a root password needs to be setup.

    >cd \mysql\bin
    >mysqladmin -u root password YOUR_PASSWORD
    

2.2.2 Download GifiArray Analyzer

• Gifi Array Analyzer for Windows 2000/XP can be downloaded from here
• Gifi Array Analyzer for Redhat Linux can be downloaded from here

2.2.3 Installation

• Download the distribution file to a directory of your choice and unzip it. You will find a subdirectory named 'GifiArray_Analyzer' created in your installation directory.
• Switch to directory 'GifiArray_Analyzer' and run Gifi_Install_xxx.bat.
• You will be prompt for the root password to your MySQL server.

3. Running GifiAray Analyzer

3.1 Input Data Format

GifiArray Analyzer takes 3 types of input data files, cDNA Microarray, Affymetrix and gene group data file. There are sample data files under director GifiArray_Analyzer/data. Your input data file should be tab-delimited text file. (Missing data is ok.)

• cDNA Microarray data file format

  Accession	EXP1	EXP2	EXP3	EXP4	EXP5	...
  Gene1	data	data	data	data	data	...
  Gene2	data	data		data	data	...
  Gene3	data	data	data	data		...
  Gene4		data	data	data	data	...
  Gene5	data	data	data	data	data	...
  ...	...	...	...	...	...	...

  The 1st column is Genebank accession number or yeast systematic ORF name.

• Affymetrix data file format

  Probe set	Accession	EXP1	EXP1 call	EXP2	EXP2 call	...
  AFFX-..._at	Gene1	data	data	data	data	...
  AFFX-..._at	Gene2	data	data	data	data	...
  AFFX-..._at	Gene3	data	data	data	data	...
  AFFX-..._at	Gene4	data	data	data	data	...
  AFFX-..._at	Gene5	data	data	data	data	...
  ...	...	...	...	...	...	...

  The 1st column is Probe set, the 2nd column is Genebank accession number or yeast systematic ORF name, the 'call' column contains one of the three characters, A(Absent), P(Present) and M(Marginal).

• Group Gene data file

  A Gene group file is a 2-column, tab-delimited text file. The first column is the group id, and the second column is gene name. A group id can be a number or a letter. Genes in the same group should stay together, for example:

  Correct	Wrong
  0 YGR214W 0 YGR148C 1 YDR418W 1 YLR388W 2 YGL031C 2 YGL123W	0 YGR214W 1 YDR418W 0 YGR148C 2 YGL031C 1 YLR388W 2 YGL123W

3.2 Walk-through using sample input data

GifiArray Analyzer should be run under the directory 'GifiArray_Analyzer'.

3.2.1 Basic Analysis

• Double click on Gifi_RUN_xxx.bat, you should see a window popped up on your screen like this:

  

• Click on Start, choose Load microarray data and you will see another window popped up for you to select your data file and enter relevant criteria.

  

• Click Open Microarray File, select GifiArray_Analyzer\data\cDNA.txt, select organism Yeast, in Ratio Threshold type in 1.5, keep eveything else the default value, and click on OK. For more information about threashold setting , please refer to 3.2.3 Threashold Setting It will take a few minutes, depend on the speed of your CPU, for the following window to be shown.

  

• Clicking on Show/hide index will allow you to see the indexes of the functional categories. Keep clicking on this button, you will see alternative names of the categories.

  

  

• Click on button Show/hide Gene will show/hide the genes in the functional categories on the graph. The genes are identified by smaller dots than the category dots. A gene dot may represent one gene or a group of genes that have the same response pattern (i.e. similar expression profile, in this case, it is actually a kind of gene clustering).

  

• When you see the Analyze button is checked, you are in the Analyze mode, which means more detailed information about a category or gene can be viewed. Click on a certain category dot, the category and the genes in that category will be highlighted on the graph. At the same time, details about that category will be shown in the main window.

  

  All the categories that contain genes that also belong to the clicked category will be listed in the left table. Information includes: the category size; size of the intersection set (i.e. contains how many genes that also belong to the clicked category); P-value calculated by using hypergeometric distribution, which is used to check if (and to what extend) a category is enriched by the genes belong to the clicked category. Calculating P-value is time-consuming, please wait. The functional information of the genes that belong to the clicked category will be displayed on the right side of the main window.

  

• Click a category in the left table, you will see the right table divided into two tables, the upper table contains genes that belong the intersection set of the two clicked categories.

  

• Click a gene in the right table, Internet Explore in Windows or Mozilla in Redhat Linux will be launched (if failed, check if your IE is installed under "C:\Program Files\Internet Explorer" in Windows) and directly go to a web database (GeneCards for Human genes, LocusLink for Mouse genes, SGD for Yeast genes) to search information of the clicked gene.

  

• Often times, the category dots are very close to each other and their indexes are overlapped. You can drag the indexes away to make the graph more legible using the right key on your mouse. Right click on the dot or the index, hold the mouse key, release it at a desired location. An arrow will be shown to connect the dot and the dragged index.

  

• Click the gene search button, input the Genebank accession number, or gene name or Unique Gene ID (case sensitive). The dot that represent the selected gene will be highlighted in the graph.

  

• Click on a gene dot to view the group of genes with the same response pattern in the main window.

  

• Click the Sel/Del Categories, you will be in the Delete mode. You can click on the category points you want to remove, the clicked point will change to pink color. Mouse drag can be used to select more than one point at a time. After click your mouse at some place and release at some other place, the points outside the rectangular area covered by these two spots will be deleted. In other words, your mouse drag will keep the points within the dragging area and delete all others.

  

  You can press Undelete Categories to retrieve all the deleted categories. To undelete a single category, you just need to click on that point again. Retrieved points are showing their original color instead of pink. Click plot->update plot in the main menu, to see an updated plot. If you want to go back to the original plot, click plot-> get original plot in the main menu.

3.2.2 Gene Ontology/MIPS Analysis

• After getting the plot, you can do Gene Ontology analysis by clicking Database->Gene Ontology Analysis in the main menu. The program will run a SQL query and display a new window containing Gene Ontology information including Gene Ontology IDs, Gene Ontology Terms and how many genes in your dataset (after filtering) belongs to each category.

  

• This Gene Ontology Information table is sortable by table column. To sort the table, you need to click on the header cell of the column that you want to sort by. The sorting order can be changed by clicking on the header cells more than once.

• Now you can select the functional categories and add them to the graph by using 'Shift/Crtl' and mouse click. When you finished your selection, go to the main menu, click on Plot-> Add Functional Categories. New points representing Gene Ontology categories will be added to the graph.

  

• MIPS Analysis is similar to Gene Ontology Analysis. Click on Database->MIPS Analysis from the main menu. The rest steps are the same as in Gene Ontology Analysis.

3.2.3 Threashold setting:

The input data are supposed to be log-ratios.

• The default ratio threshold (for selecting differentially expressed genes) is 1(=log2(2.0)).
• Row threshold (r_th) is used to filter genes that are not differentially expressed in at least r_th samples. The default value is 2.
• Column threshold (c_th) is used to filter categories that contain too few genes. The default value is 20 (e.g. if a category contains less than 20 genes it will not be included in the analysis).
• Weight is the relative proportion of expression categories and functional categories, the range is (0, 1). If you do not want the structure of the data to be effected dramatically by adding functional categories, you should use large weight (for example 0.9).

3.2.4 Other features:

• Zoom: Graph can be zoomed in and out on both X and Y axis;
• Print: Tables and graph can be printed directly to your default printer;
• Data Export: Graph image and table content can be saved to your local hard disk;

4. How to cite Gifi Array Analyzer

5. Contact information

• For techinical questions regarding system installation, execution and bug reports please email to Mi Zhou at University of Tennesee Health Science Center

• Please address all other questions and comments to: Yan Cui at University of Tennessee Health Science Center, or Wing Wong at Harvard University.