1. Introduction :

This java applet displays the network extracted from Pathways present in the three public Pathway resources. To know in detail about how the networks are extracted click on the Help button on the top bar or here. The applet runs using Java runtime environment (JRE 1.4.2) and if this is not installed on your machine, the program will detect it and install it on the machine. This is only one time install and from then on it will run every time on its own. Once the applet starts to load a certificate will appear for to accept the applet and grant the session. This can be accepted for one time session or for all subsequent sessions on the machine. This is being done to get the permission so that the images and text can be saved on to the machine if requested by the user.

 


Once the session is granted a dialog box will appear while the extraction network is downloaded. Depending upon the speed of network connection and the memory available to the applet this might take a while to load. For heavy network this might take longer. This can be improved to some extent - see under Problems. By default, for the Pathway type (metabolic or cellular regulatory) that the user has selected, the graph network with maximum nodes and interactions appears first in the window. This can be from any of the resource. Further options under menu bar are available to rebuild the networks from other resources or filter it based on user criteria.
 
2. Menus

The menu bar has many features for visualizing and analyzing the network and the microarray data sets in context of the gene associations in various pathways. The main buttons that show on the menu bars are identified by numbers on the Figure and are as following:


1. Drop down menu for different experimental data for the network. For details see 2.1 Expression values from multiple samples or experiments

2. Drop down menu for filtering network based on various criteria. For details see Filters 2.2.

3. Selection for Pathway type - Metabolic or Regulatory cellular processes from any one of the three resources. For details see 2.3. Pathway type

4. Select network extracted from genes from both metabolic and regulatory pathways from one of the resource. For details see 2.4. Pathway Source

5. Select and browse nodes from selected Pathway or Gene Ontology terms. For details see 2.5. Select.

6. Save the graph as image and the node-edge information as text.

7. Drop down for browsing the Sub networks

8. Find a Gene in the network by gene name or Locus ID. For details see 2.8 Find a Gene in Network

9. Refresh data For details see 2.9 Refresh Data

10. Resource and the type of Pathway chosen.

11. No. of accessions in user given dataset that were mapped to Pathway resource and the experiment Number if expression values are supplied along with the accessions. For details see 2.1 Expression values from multiple samples or experiments.

12. The number of subgraphs for the given genes. All sub networks found can be browsed using the drop down subgraph menu (7).

13. Current sub network shown in the applet window.

14. Information about nodes (genes-Locus Ids) and their connections (Pathways)

15. Go Back to the previous graph.
 
 
 
2.1. Expression values from multiple samples or experiments

If the experimental values are not included in the dataset then the graph network appears without the experiment menu button and appears like the figure below. The nodes having the color grey.




Each column of log ratio is treated as an independent experiment. If multiple columns of experimental values are provided along with the gene ids then the Experiments button appears on the menu bar. This drop down menu will have the experiments listed based on the no. of column of log ratios that are submitted. This is in the order the user submits the log ratio. The first column being Experiment 1 etc. This feature helps to compare and evaluate the node expressions for the same network for different experimental data set : say data from different time points or data after any differential treatments. Alternating between experiments helps in visualizing and analyzing the changes in gene expression for all the associating pathways under differential conditions. The left panel in the applet shows the Experiment Number for the current network. By default when the graph loads initially this is number 1, the first column of log ratios that the user supplies.




 
2.2. Filters

There are three different types of filtering option available:

2.2.a. Node Connections

Node Connection filter is an option that can filter the network based on user given connection strength. The weight to the edges is given based on the number of pathways in which the two participating nodes coexist together. There are four different edge strengths that are given. This is for visualizing and finding the differentially regulated genes that together affect multiple pathways.
 




 
2.2.b Expression

Alternating between different experiments, networks can be filtered for up regulated and down regulated genes in the given network. The Expression filter under Filter menu provides two options - Up and down. This can let the user evaluate which genes and pathways are activated or repressed in different datasets. Combined with alternating between experiments for the same network up regulation and down regulation feature can highlight the most differentially regulated genes and in turn the pathways.




 
2.2.c. Combination Filter

A combination filter combines the two independent options under the Filter menu. This filter can be used for visualizing genes that are either up regulated or down regulated or both and genes that participate in more than one pathway (have one or more edge strength).




 
2.3. Pathway Types

Visualizing only Metabolic or cellular regulatory pathways. Graph networks can be chosen for visualization from any one of the three resources using the Pathway type menu.




 
2.4. Pathway Source

Networks can be build up from both metabolic and cellular regulatory pathways from any one given resource choosing one of the Pathway resource menu options. This helps in identifying genes that might be playing a role in both type of pathways and possible connections if any between two types of biological pathways under any differential conditions that are being tested.




 

 
 
2.5. Select

2.5.a. Nodes belonging to Pathways

Within the dense networks that are built up using all the genes in a given user dataset, nodes (genes) that correspond to specific pathways of interest can be chosen. Choosing the pathway option, a list of pathways that have built up the current network is presented. One pathway or more than one pathway can be selected by holding down the control key. The selection will highlight only the nodes and the interactions that belong to the user chosen pathways.




 
2.5.a.1. Pathways with significant P-value

If statistical analysis is chosen as an option then an additional feature appears under the selection menu. Select pathways based on their P-value. Four options are provided. After selection, the pathways that pass the criteria will be highlighted as red. The edges that did not pass the criteria and their nodes will be reflected in grey. Nodes that are associated with more than one pathway and not all of them pass the criteria appear as black.Clicking on the edge will take to the pathways.




 
2.5.b. Nodes belonging to Go terms

This is an additional option for selecting nodes of interest. The underlying database that holds the pathway information also holds the information on Gene Ontology terms that are mapped onto the genes. Selecting this options genes that belong to one or more GO terms can be looked up. Choosing one of the ontologies- BP, CC or MF will bring up all the GO terms that have been found to be represented in the dataset from which the network has been extracted. One or more Go terms can be chosen by holding down the control key, this will highlight the genes that are mapped to the Go terms.




2.5.c. Nodes with two or more neighbors

Nodes which have more than one neighbor can be selected from dense network using the multiple options provided under this drop down menu. This is helpful in analyzing the genes that have multiple associating partners in one or more pathways. The highlighted nodes are the ones that pass the chosen criteria and can be further investigated by options provided in Section 3.5 and Section 4.




 
2.5.d. Node that has maximum number of Neighbors

Making this selection will highlight the node that has maximum number of neighbors. This is the gene that has most associations with other genes in the user given dataset. These associations all could be part of one pathway or multiple pathway.




 
2.6. Sub graphs

The graph networks that are built for any given dataset from any given resource for any type of pathway may result in more than one sub networks based on the associations of genes in pathways. These sub networks can be visualized using the Sub graph menu option provided in the main menu bar.




 
2.7. Save

Chosen network on the screen can be saved as an image using the save option in the menu. The Node (genes)- Edge (Pathway) information for the current network that appears in the left panel can be save as text file.
2.8. Find a gene in a Network

The node edge info appears in the left panel shows which genes are associating with which other gene in any pathway. The Nodes in the network have locus IDs and the gene names as identifiers. To locate a gene in the network, click on the find button, this will bring up a dialog box where you can either enter a gene name or Locus ID. This will highlight the node and also its connections with neighboring genes.
2.9. Refresh data

After applying filters, to revert back to original graph use the refresh data button.
3. Network features

3.1.Zoom in

3.2.Zoom out

3.3.Scale to fit

These are the standard features of any applet. These can be applied from the menu bar or by right clicking on the node. This brings up a menu which has all these features.

3.4.Drag the nodes

In dense networks the nodes can be dragged all the way out of the network for better visualization and understanding. By selecting a node and holding down the left button of mouse, nodes can be dragged around.
 
3.5.Select Neighbors for one Node or multiple nodes

To see the connections in a highly dense networks right click on a node and select Show neighbor option from the menu. This will highlight all the neighbors and their connections to the node. To select multiple nodes hold down the control key and after selecting right click on any node and select the "Show neighbor" option.




3.6.Deleting Nodes

To delete a node and its connections from the network right click on a node and select Preview deletion from the menu. This will show a preview of all the connections that will be removed if that nodes is deleted. To go on and delete it right click and select "Perform deletion". To delete multiple nodes hold down the control key and after selecting right click on any node and repeat the same.




4.Node information

Selecting a node and double clicking it on it will take to an Information page that presents a detailed report about the selected gene. This contains data compiled from multiple sources and has links to external databases.




4. Edge information

Selecting a edge and double clicking it on it will take to an Information page that presents the list of pathways in which the two nodes co-occur. Clicking on the pathway name will take you to the Pathway map. The map contains all the genes highlighted that are present in the user supplied dataset..




6.Problems in running the Applet:

1. Applet not responding:
The reason could be that the dataset you have provided might have a large number of interactions. There are two time consuming steps in the graph building process, one is to get the data from the server and the second is to draw the graph. Both have been optimized to give faster response but it varies from dataset to dataset. A message window appears while loading the graph, wait till that window disappears.

2. Applet might be needing lots of memory
to do certain operation which is not available. When an applet runs on your machine by default there is an upper limit set on the amount of memory the applet can use , this will be typically around 96 MB. Thus it may be possible that when you are working with a huge interaction network, the response from the applet starts becoming sluggish. This is because all the memory is being used for building the interaction network and little is left behind for doing further operations on the network.

If you want better performance from the interaction network displaying applet when you are dealing with hundreds of nodes and edges then you can increase the memory allocated to the applet resulting in a better performance.

Follow these steps to do this.

1. Close all the browser windows.

2. Click on the Start Button , and then on the Settings , followed by Control Panel




3. Open the Control Panel and look out for the Java plug in icon.
You may have multiple Java Plugin icons on your Control Panel , Select the one with Version number being greater than equal to 1.4.0 .

For determining this click on the Java Plugin Icon and click on the About Tab and note the Version. If the Version is equal to greater than 1.4.0 then proceed else click on another Java Plugin icon.
4. To increase the amount of memory allocated to the applet , click on the Advanced tab ,
where you will get a window like below .



Now in the text field titled as Java Runtime Parameters, fill in as -Xmx128m And click on the Apply button. Here you are allocating 128 MB of memory for the applet , and the applet would be able to respond quickly when you are dealing with heavy networks or intense operations on the network. More memory could be allocated in similar fashion, by replacing 128 MB to 256 MB. But it might not be needed since applet should perform efficiently at 128 MB.



 

Citing Biorag : If you have used or are using BioRag for your research please cite or acknowledge this resource in your publications as "BioRag (Bioresource for array genes) at www.biorag.org".

For any comments or questions contact Dr. Ritu Pandey or Prof. David Mount.
BioRag database is maintained by the Bioinformatics group at Arizona Cancer Center. The material presented here is compiled from different public databases. BioRag is hosted by the Biotechnology Computing Facility of the University of Arizona.© 2002,2003 University of Arizona. All Rights Reserved.