|Gifi Array Analyzer|
Gifi Array Analyzer is a Java application for integrated analysis of DNA microarray data and gene functional classification data.
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.
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.
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.
Windows 2000/XPcan be downloaded from here
Redhat Linuxcan be downloaded from here
Gifi Array Analyzer is written in Java and has been tested on Windows 2000/XP and Redhat Linux 8.0.
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
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'.
>cd \mysql\bin >mysqladmin -u root password YOUR_PASSWORD
Windows 2000/XPcan be downloaded from here
Redhat Linuxcan be downloaded from here
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.)
The 1st column is Genebank accession number or yeast systematic ORF name.
|Probe set||Accession||EXP1||EXP1 call||EXP2||EXP2 call||...|
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).
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:
GifiArray Analyzer should be run under the directory 'GifiArray_Analyzer'.
Gifi_RUN_xxx.bat, you should see a window popped up on your screen like this:
Load microarray dataand you will see another window popped up for you to select your data file and enter relevant criteria.
Open Microarray File, select
GifiArray_Analyzer\data\cDNA.txt, select organism
Ratio Thresholdtype 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.
Show/hide indexwill allow you to see the indexes of the functional categories. Keep clicking on this button, you will see alternative names of the categories.
Show/hide Genewill 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).
Analyzebutton is checked, you are in the
Analyzemode, 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.
drag the indexes awayto 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.
gene searchbutton, 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 dotto view the group of genes with the same response pattern in the main window.
Sel/Del Categories, you will be in the
Deletemode. 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 Categoriesto retrieve all the deleted categories. To
undeletea single category, you just need to click on that point again. Retrieved points are showing their original color instead of pink. Click
plot->update plotin the main menu, to see an updated plot. If you want to go back to the original plot, click
plot-> get original plotin the main menu.
Database->Gene Ontology Analysisin 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.
The input data are supposed to be
ratio threshold(for selecting differentially expressed genes) is
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).
Weightis 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).
Zoom: Graph can be zoomed in and out on both X and Y axis;
Data Export: Graph image and table content can be saved to your local hard disk;
For techinical questions regarding system installation, execution and bug reports please email to Mi Zhou at University of Tennesee Health Science Center
Term of Use: Gifi Array Analyzer can be freely
downloaded for academic, non-commercial use only.
This software is provided "AS IS". There is no warranties,
express or implied on it.
Thank you for using Gifi Array Analyzer.
|Last Updated: April 18 2004||Report website problem to: Mi Zhou|