RPGeNet v2.0

Tutorial


Table of Contents

  1. Network Explorer
  2. Pathway to Level
  3. Shortest Path Between Genes
  4. Guided Tutorial with CERKL

Network Explorer


The Network Explorer allows you to explore the different levels of the Retinitis Pigmentosa (RP) network by selecting one or several genes and retrieving all their interactions.


Form options



  1. Gene/s.
  2. Here is where you can input one or several genes of interest related or unrelated to Retinitis Pigmentosa (RP) (althought if is not related there is a big change that it won't have interactors, since the network is using RP genes as seeds). If the input are multiple genes, then those genes have to be separated be a comma "," or a blank space " " between them.

  3. Interaction Level.
  4. The RP network has 4 levels: "Skeleton", Level 1, Level 2, Level 3, and WholeGraph. Each level represents a sub-network of genes and interactions ranging from more related to Retinitis Pigmentosa (Skeleton) to less related (WholeGraph).

  5. Gene distance.
  6. Specify the distance from the input genes to retrieve from the Network. By default this is set to 1, that is, RPGeNet will only retrieve parents and childs of the specified genes. If set to 2, RPGeNet will also retrieve parents and childs to the interactors of the input genes. Please, note that this field can only be changed when Interaction Level is set to Skeleton.

  7. Expression.
  8. Color the genes in the graph according to their expression in the specified experiment.



Navigation Options

In Network Explorer you have several options and controls to modify the graph visualization.

  1. Home:
  2. The logo image is a button to return to the home page.

  3. On Click:
  4. This option defines three different behaviours when clicking on Nodes or Edges. "Show Properties" will display an overlay with information about the gene or the interaction. "Node Addition" will add all the interactors to the clicked node to the visualization. Finally, "Node Deletion" will delete the clicked node from the visualization; please, not that this process is irreversible.

  5. Layout Selection:
  6. Change the layout of the nodes and interactions. Available layouts are: 'Dagre', 'Cose', 'Grid', 'Circle', and 'Concentric'.

  7. Connect Genes:
  8. This button will add all the interactions in the current level between all the genes displayed on the visualization.

  9. Remove Unconnected Genes:
  10. This button will remove all the nodes that are currently not connected in the graph display.

  11. Border size:
  12. Increase or decrease the border size of genes.

  13. Node size:
  14. Increase or decrease the border size of genes.

  15. More Controls:
  16. Legend and Tutorial:

Legend

The roots of the RPGeNet network begin with the driver genes of retinitis pigmentosa. All driver genes are purple borders. Some driver genes are associated with syndromic genetic disorders and are square shaped within the network explorer. Non-syndromic driver genes are diamond shaped. Driver genes that are syndromic and non-syndromic are star shaped.


Non-driver nodes are all circular and have borders that indicate the level, where the gene is found within the network, by the color.

The arrows that connect nodes within the network are color-coded to indicate whether they are physical, genetic or unknown interactions.



All nodes, driver and non-driver genes, have a colored interior that represents their expression depending on the expression data used. Absolute expression is the raw expression within the retina. The RET-ALL option is the relative expression within the retina in comparison with all 32 tissues of the microarray experiment. The RET-OTHER is the relative expression within the retina in comparison with all the tissues minus the retina.



Pathway to Level



Pathway to Level allows you to get the shortest paths between a specified Gene and either 'Driver Genes' or 'Skeleton Genes'. If there is more than one target gene with the same path length (the same number of interactions until reaching that gene), all the pathways will be displayed on the results page.

Pathway to level is useful when trying to assess the importance of a particular gene to Retinitis Pigmentosa, or to have an idea of how close a gene is to other genes related to Retinitis Pigmentosa (in terms of signalling and protein pathways).


As can be seen in the image above, each pathway is represented as an individual graph visualization. Only one pathway is represented between two genes; if you want all the possible shortest pathways between, for example, PMM2 and SNRNP200, you can click on the "All shortest paths" link above the visualization. In addition, if you want to explore a particular pathway further, you can click on the "Explore Network" button, which will bring you to Network Explorer with the specified pathway in the graph visualization.



Shortest Path Between Genes



This form allows you to compute the shortest path between any two genes in the network. Please, note that all the connections in RPGeNet are directional, so the results of A->B will be different from B->A. All the shortest paths are computed on the WholeGraph. The results page from this form is the same as the one from "Pathway to Level".

Guided Tutorial with CERKL


This guided tutorial will show you how to:

  1. find your gene of interest in the network
  2. learn how to edit, expand or contract a graph in network explorer
  3. learn to set and change the expression visualized
  4. learn to find the shortest paths between two genes of interest
  5. learn to find all the pathways between your genes of interest to drivers/skeleton

Starting from the RPGeNet homepage we see three options available to users so you can use: Network Explorer, Pathway to Level and Shortest Paths. Below we will go through each on of these options and guide you through the process.

Network Explorer

The first option available is the Network Explorer. This option is great for when you have a gene of interest and want to know what reacts with your gene, when you want to create a graph of the interactions with your gene and when you want to explore network with your gene of interest as the focal starting point. If we take a look at Network Explorer we are asked to type in our gene(s) of interest, choose the interaction level, the gene distance and the expression dataset.

In the gene section, we will type in the gene CERKL but do note that you can type in multiple genes and not only one.

The interaction level is basically asking us to chose a level of penetration into the network you wish to search through and how complex of a graph you want returned. The most basic level in the options is the skeleton. The skeleton is composed of all the shortest pathways between the driver genes of retinitis pigmentosa and are the most relevant interactions related to retinitis pigmentosa. If we pick the skeleton, Network Explorer will only search through the interactions of the skeleton graph and the output will only include interactions, of your gene of interest, found in the skeleton. Each level above the skeleton adds the parent and children nodes of the previous level that do not exist within the previous level(s) of the network. Depending on the level you choose, you will have more genes, interactions and pathways to search through but also an increase in noise as you move further away to interactions not as vital to retinitis pigmentosa.

For the purpose of this guided tutorial, we will keep it simple and chose skeleton.

The gene distance option is only available if the skeleton is chosen as the interaction level. This option is simply there to save time and automatically add the parents and children of all the genes that interact with our gene of interest in the skeleton. Note that only parents/children found within the skeleton will be added if there are any. We will keep the defaul of one.

Finally, we must choose the expression data. There are four options,all from the same microarray dataset, GSE7905, that have been uniquely calculated. The first option is the absolute expression values of genes within the retina only. The second option is the relative expression between retina and all the 32 tissues analyzed. The third option is the relative expression values between retina and all non-retinal tissues. Lastly, the final option is the relative expression between retina and the liver.

We will chose the absolute expression option and then click on explore network.

The image above is the result of our input: a graph with our gene of interest as the focal point and all genes that interact with it in the skeleton. You can also see there is a control panel to the righ hand side and other tools you can use to help you navigate the network and find what you are looking for.

Lets take a look at the control panel to the right hand side of the screen.

The control panel offers helpful tools to help navigate the network, edit your graph and save the graph as an image or for later use. We will go through each button starting from the top. The first section, called "On click", has three buttons and affects the output of your click when clicking on nodes of your graph.

The first button, "Show Properties", is the default setting and is used to find more information about a gene or an interaction. When on "Show Properties", all you need to do is click a gene that interest you and a popup window will open.You will find helpful information like its alias, a summary about the gene, gene data, expression data, its gene ontology and links to other well known gene databases for more information. The image below is the result of clicking on the gene VHL, when using "Show Properties".

We can play with our graph and add or delete nodes. If we click on the node addition button and then on a gene that interests us, we can add the parents and child of that gene. Go ahead and click on "Node Addition" and then click the genes PPM1A and PPM1B.

If there is a gene that does not interest you, you can click on node deletion and double click on that gene to delete it. If there are many genes you want to delete, you can press and hold shift and draw a highlighting box around those genes and then click on one of the genes highlighted to delete them all.

There is also an option to change the layout of the graph. There are ten different layout options you can pick from. Go ahead and pick one by one to see which one you like the most

When we add new nodes to the graph they only show their interaction with the gene you clicked. If you want to see how the added node connects with the rest of the graph, you can click "Connect Genes". You can even click the same button before adding nodes to see if the genes that interact directly with you gene of interest interact with each other.

When you delete nodes, you may end up with many unconnected nodes in the graph display. We can use the Remove Unconnected button to remove any unconnected nodes. Restarting with the original CERKL input on the netexplorer, we are going to click on "Node Addition" and click on the gene SIRT1. We will then click on "Node Deletion" and delete SIRT1. All of nodes connecting to SIRT1 will be unconnected. Now click "Remove Unconnected" and all the unconnected nodes should be deleted from the graph display.

As you can see, each node has a border that tells us the lowest level the gene is found. The size of the border can be modified using "Border Size".

You may have noticed that the nodes have different sizes. The size corresponds to the number of known variants. If you would like to turn off the node size you can click "Disable" under "Node Size" on the right hand ....

Under "Controls" there are many useful features you can use. The "Fit" button resizes your graph so that it fits on screen. If you want an image of your graph, you can click on the "Image" button and a PNG image of your graph will be created for you to download. If you want to a list of the interactions in your graph, you can click the "Export" button that and a TBL file will be created and can then be opened on Excel.

If you would like to save your current graph for later use, you can click "Save" and a JSON file will be created for you to download. Give it a try and click "Save".This JSON file can then be uploaded to RPGeNet. If we restart and go back to the Homepage, we can click on "Explore Network" without typing any gene.You will get a "No Results" warning but can click on "Upload Graph" to upload the JSON file.

You can also retype CERKL and get the standard graph and then upload using the "Upload" button under "Controls".

If you need help or are lost, the legend and tutorial buttons are available, found at the bottom of the control panel.

You will notice that the nodes have different shapes and colors as do the interactions. We can quickly check what the colors and shapes mean by looking at the legend, that is found scrolling down the control panel on the right.

If you decide, whiles exploring the network or fixing your graph, that you want to see different expression values, you can always click under "Expression", on the top right hand side. All four options are available to chose from. Currently we have "ABSOLUTE" expression in the retina. Why don't we change it to relative expression "RET-ALL".

You can see that the color on the inside of the nodes changes. The absolute option has a different coloring scheme than the relative options. You can click on "Show Legend" for the color scale.

Pathway to Level

The second option available is Pathway to Level. This option is used when you want to find all the pathways that connect your gene of interest to the driver genes of retinitis pigmentosa or to the skeleton level of the network. This is useful for genes that may not be a driver or part of the skeleton and you want to predict how relevant it may be.

You can type in a gene of interest or try out the gene "RAX". In the "Path to" section, you must chose between finding a pathway to the skeleton or to driver genes.

We will keep the default, drivers, for the "Path to" section. Like in Network Explorer, you also have to chose the expression data. We will also keep the default, ABSOLUTE, and click "Get Path to Level".

The output should be a list of pathways between the gene inserted and drivers if the pathways exist. In the case of RAX, we have three pathways.

When you find a pathway that interests you, you can transfer the pathway to Network Explorer by clicking "Explore Network" on the top right of the pathway. The output should be your chosen pathway on Network Explorer. Here you can use all the tools we used in the Network Explorer tutorial (above).

Shortest Path Between Genes

The final option is the Shortest Path Between Genes. This option is great for finding all the pathways between two genes of interest. This time you have two gene inputs: "Source gene" and "Target gene". Lets use CERKL as our source gene and RHO as our target gene. Do note that Shortest Paths is directional and so you need to type the genes in order from upstream (source) to downstream (target). If you want to see the opposite order, you must restart the query and switch the order of the genes. Like in the previous tools, you were asked to pick the expression data. You have the same four options and can choose the one best for you or keep the default for this tutorial. Than click "Get Shortest Paths".

Like in Pathway to Level, we get a list of pathways, if they exist. In the CERKL to RHO example, there are three shortest pathways. We can transfer the pathway that interest us to Network Explorer by clicking "Exlore Network".


This concludes our guided tutorial of RPGeNet.