Pathways, interactions and entities, oh my!

Our website is based on an integrated database (MetNetDB) with integrated metabolic and regulatory pathway information. Initial data comes from a variety of sources, and has been manually curated and enriched. The database however was lacking a userfriendly low-threshold interface to the public (as well as our own on-site researchers). That why we built the website that you see on the left.
Probably the first thing you're curious about is what we content we actually have. In the header of the page you see a two-level menubar. When you click on the Browse heading, you get taken to the first option: Browse pathways. This presents you with a split screen. On the left-side is an ontology tree. On the right you can see the pathways that are in your currently selected ontology (default is the top-level of the ontology, so all pathways are shown).. Seeing all pathways in a single list is a bit overwhelming (300+ on the first visit, and possibly even more on subsequent visits depending on your settings). The ontology that you see left side helps to classify pathways. The ontology is hierarchical.
Before we go into the tree, it is worth pointing out that when you see a lot of pathways in the right panel, you can toggle the way that they are displayed. In the top-right corner of the panel you see two buttons. We highlighted them in yellow for you in the screenshot on the left. The first button is to display the pathways in a list. The second button reverts to showing thumbnail representations of the pathways. Either way you see the same pathways, it just depends how you prefer to have them displayed. Any time you see this two-panel pathway representation, you will have these buttons at your disposal to toggle between representation modes.
You can expand and collapse branches in the ontology tree in the left panel by clicking on the and buttons in front of each heading. In the screenshot, we've expanded the Biosynthesis and interconversions category.
In order to get an idea of how many pathways are in a category (or if it's worth going into), you can hover over any item in the ontology. A tooltip-popup will tell you how many pathways are in that particular category.
Once decided, you can click on the category-link to change the content in the right panel. The panel will then only display the pathways in the category that you selected. In the screenshot, we picked Carbohydrate biosynthesis.
Pathways can be previewed as well, even if you are in list view mode (as apposed to thumbnail view mode). However the mouse cursor over the name of the pathway and you will see a popup-window with a smaller-scale representation of the pathway. In the screenshot, we're hovering over the Gluconeogenesis pathway.
After clicking on name of your pathway of choice, you see more details of the pathway. It usually starts out with some comments on how the pathway constructed. Literature references in the text can be clicked on when there's a icon. This will take you to the PubMed abstract page for a particular reference. For the pathway commentary, you also get to see who made the comment. If it says "AraCyc" or another well-recognized pathway database, we imported the commentary from that original source. Otherwise, it may be comment that was manually added by one of our own internal curators. Next to the textual description and commentary, a thumbnail of the pathways is displayed.
As you scroll further down the page, you see more detailed information. You get an overview of all the subcellular locations (organelles) where the pathway participates. You can look at the interactions that the pathways is made of. Below that, you see the (molecular) entities that participate in the pathway. All of these are grouped by different categories, because some pathway may contain a lot of them. You may not even be interested in seeing all of them.
Expanding the different categories works in the same way as the expansion of the ontology-tree that we did earlier. The icon expands, the icon collapses. In the screenshot we expanded the Enzymatic reaction section of the pathway interactions.
You can do the same for the Entities that make up the pathway. Here we expand to look at the metabolites.
Back to the top of the page, you should see a thumbnail of the pathway next to its description. When you click on the thumbnail, you will be taken to a new page that offers visual interface to the pathway.
This is the visual interface to pathways and networks. The interface is somewhat similar to Google Maps. At the very top of the browser viewing area, a basic toolbar remains visible to switch to other parts of the site. Underneath the toolbar, several frames are visible. The top left frame shows a thumbnail view of the pathway, with a green cursor overlaying the part of the pathway that's currently magnified in the (largest) frame on the right. Underneath the thumbnail frame, another frame provides links to the different components of the pathway, including a link to go back to the pathway's textual description.
Let's go through some of the details in the main frame where the pathway is shown in detail. On the right-hand size in this screen, you see a legend explaining the different symbols and coloring schemes we use for our pathways. You can "hide" the legend to see more of the pathway (this behavior has been reported to not work on all webbrowsers). There's smaller toolbar here that lets you resize the pathway to fit more or less of it on your screen. In the screenshot, we selected 66% of the original resolution, a format which is still very well readable on most screens. Note that the size and the position of the green cursor in the thumbnail frame automatically adjusts as well.
In this screenshot, we reduced the pathway to 33% of its original size. This is probably too small on most screens and certainly the text of the individual nodes is no longer readable. However, this can still be useful for a structural examination and a quick inspection for such features as cycles.
You can customize the size of the presented image by clicking on the "Custom" link in the resizing toolbar. You can pick any number between 0.01 and 1.00, whereby 0.01 means you wish to see the pathway scaled to 1% of its original size, and 1.00 means you wish to go back to the original size (100%). In the screenshot, the entered value of .50 means we want to have the pathway size to 50% of its original size.
This is the result of a 50% reduction of the original size of the pathway.
A cautionary note with regard to sizing: Custom sizing has to be done on demand, whereas it is quite possible that one of the standard .66 or .33 factors was already selected by somebody else before you. In that case, the resized image is cached on our server and will load quicker than any custom resizing you decide to do yourself.
There are two ways to move the pathway around. You can drag and drop the cursor around in the thumbnail frame, or you can grab on to one of the nodes directly in the main frame and drag the pathway around until you find what you're looking for. If you do the latter, you'll see the cursor dynamically moving along, so you're never really lost thanks to this thumbnail map.
We returned to the pathway by clicking its link in the information frame underneath the thumbnail frame. Note the different tabs in this view. So far we've only talked about the "General info" one, but there's much more to explore!
Click on the "Linked pathways" tab. Pathways are arbitrary textbook concepts, and the output of one pathway can certainly be used as the input for another one. A classifical example is pyruvate: considered an end-result of the glycolysis pathway, it's also input for the TCA cycle.
So with "Linked pathways", you can see for each pathway how it connects and interfaces with other pathways. The first heading presents a list based on comment genes, RNA, proteins and protein complexes.
Since our selected pathway is gluconeogenesis, it should come as no surprise that it shares features with other sugar-related pathways, including glycolysis and the TCA cycle.
A second paragraph describes metabolites in the pathway that can also be found in other pathways. In contrast with genes and other entities, the actual pathways are not shown. This is because there are many common small metabolites that participate in many pathways. Think of ubiquitous compounds such as O2, ATP and even our earlier example pyruvate. So, for linking metabolites, it is left to the visitor to select specific metabolites that are of interest and further explore them.
A third level of pathway linkage is possible: two pathways can be related if each contains a protein that interact with each other. Protein-Protein interactions can also refer to and confirm reactions within the pathway under examination. Quite often e.g. we see complex formation in MetNet confirmed through atPID.
For visual guidance, proteins that are inside the pathway are rendered in blue. Proteins outside the pathway are rendered in red. A more detailed view of overlayed atPID information on top of a selected pathway can be seen under the next tab "Protein-protein interactions".
Protein-Protein interaction databases provide another level of potential interaction between proteins in your selected pathway as well as between the proteins in your pathway with other proteins (which in turn may or may not be part of other pathways). MetNet Online incorporates atPID for this type of interactions. The "Protein-protein interaction" tab shows an interaction network derived from atPID and based on an initial selection of proteins from your pathway.
Proteins in blue are currently in your pathway; proteins in red are not (yet) included in your pathway. A blue edge between two blue nodes means that both interacting proteins are in your pathway, and their interaction is confirmed by atPID. A red edge between a blue and a red node means that a protein in your pathway is also known to interact with a protein outside of your pathway. This allows you to form a hypothesis about whether the protein should be included in the pathway or not. In order to (dis)prove this hypothesis, a wet lab experiment has to be carried out that perturbs the behavior of the putative protein.
There are more ways to browse pathways than by ontology. One of these is by organelle. Click on any organelle in the hierarchy, and you will see a list of pathways that at least partly participate in your selection.
This selection shows 4 pathway which partially occur in the thylakoid disks of the plastid.
As a reminder: almost every time that you see a list of pathways, you can toggle between thumbnail and list view.
If pathway ontology nor subcellular location don't suite your needs, you can select pathways based on the types of interaction that take place. Examples include up-regulation and down-regulation, transport etc.
Admittedly some categories are less useful for filtering, such as transcription or translation (since virtually all pathways include these types of events).
This selection shows the presence of 22 pathways that include at least one negative regulatory event.
Finally, you can select entities by themselves. This is a useful interface if you know exactly what entity you're looking for. First, you need to select what type of entity you're interested in.
Now you select the first letter of your entity of interest. Say that we're looking for ATP. Find the letter 'A' and click on it.
Since there are more than 500 entities starting with "A", a further drill-down is presented. Go ahead and select "AT".
Now you'll find "ATP" in the list. Note the checkboxes at the top of the frame that are currently unchecked.
You can see synonyms and abbreviations for your entities by selecting the boxes at the top of the frame. While these aren't shown directly in the alphabetical listing, they are included in search requests.
Once you click on your entity of choice (in our case "ATP"), you see the detailed information sheet for your selection.
As you can see, ATP - not unexpectedly - participates in a great number of pathways.
For the visitor's convenience, we have the pathways color-coded by location. So, if ATP in a particular pathway participates in a mitochondria, the pathway we have a pink color indication in front of it. Similarly, if ATP is found in the plastid while participating in a pathway, a green indication is used.

This concludes our tutorial on how to browse the MetNet Online website. You can select another tutorial if you wish.

Copyright Wurtele lab