The {{model_organism}} Reaction Network

The {{model_organism}} Reaction Network ({{browser_name}}) has been developed with the specific aim of providing a framework for performing metabolic pathway analysis on omics data. To this end, knowledge from several online resources has been curated and integrated, and a browser has been written to inspect the pathway analysis results. Currently, {{browser_name}} is based on

The {{browser_name}} pathway browser

Metabolic pathway analysis in the {{model_organism}} Reaction Network consists of: 1.) finding all the shortest paths between the compounds that were measured on the omics platform, and 2.) visualizing the resulting subnetwork in the pathway browser. Since metabolic networks are typically densely connected due to a number of frequently occurring compounds or hubs, only paths through a reaction's primary substrates and products are considered. In addition, each path is checked for stoichiometric and thermodynamic consistency as based on the concept of Elementary Flux Modes (EFMs), and consequently represents a biochemically feasible flux in the reaction network. When selecting the "Network" tab in the browser, the resulting paths are shown on the right and the global network of connected compounds is shown on the left. Compounds are colored according to their value and are printed in boldface when their levels change with statistical significance, as shown in the legend. Intermediate compounds that were not measured on the omics platform are shown in white. Nodes and links in the network can be inspected by clicking on them, after which the corresponding paths will be shown on the right. If more than one hypothesis was tested on the omics data, the and p-values related to each hypothesis can be selected by using the dropdown menu in the upper-left corner.

The position of the nodes in the network can be changed by dragging a compound with the mouse. Subsequently the position of the node is refined automatically with respect to neighbouring nodes by a global force function; in order to switch off this effect and fix a node's position it must first be double-clicked. Panning of the network view is done by selecting and dragging an empty part in the network and zooming is done by scrolling the mouse wheel. The network view can be exported as SVG file by clicking on the menu icon and subsequently on "download SVG". Importantly, this feature only works in the Firefox and Chrome web browsers.

It is possible to exclude certain compounds from the network visualization by adding them to the exclusion list. This feature is useful to interactively exclude paths that have no biochemical significance and therefore add unnecessary complexity to the network. Depending on the platform and research question, any type of compound or group of compounds may be added to the exclusion list, such as acetyl-CoA, ATP, D-isomer amino acids, etc. For instance, when visualizing data from an acylcarnitine platform, L-carnitine connects each of the measured compounds and is best excluded from the visualization. In contrast, when analyzing small molecule metabolomics data it will be useful to show L-carnitine in the network and it should therefore not be present in the exclusion list.

The "Data Table" tab in the menu shows the table with the analyzed omics data, and contains the original metabolite name, the mapped {{browser_name}} identifier(s), and the statistics per compound. If a compound could not be found in the model it is colored red. The full {{browser_name}} model can be inspected by selecting the "Knowledge Base" tab in the menu. In the text field in the upper-right corner, metabolite, gene and enzyme names and identifiers can be searched.

The {{browser_name}} browser was implemented using the AngularJS framework and with Bootstrap style, and the D3 library was used for the network visualization.

FAQ

A compound measured on the omics platform I used is present in the {{browser_name}} model but was not mapped correctly.
Metabolites and genes are mapped by their name, including known synonyms, and by external identifiers such as ChEBI, KEGG, LipidMaps and HMDB ids for metabolites, and Entrez, Ensembl, and Uniprot ids for genes and proteins. If you have identifiers from multiple sources per compound, please provide all of them in separate columns when submitting your data. If the identifier or compound name is unknown, or the compound was mapped to the wrong entity in the model, please provide us with a list of mismatched mappings.

Some reactions in the {{browser_name}} model are lacking or have incorrect substrates or associated genes.
The content of the {{browser_name}} model is continuously updated in order to correct errors and to add missing metabolites, genes, enzymes, reactions and pathways that occur in the organism's metabolism, and your help and feedback in this process is very much appreciated! If the pathways in which you are interested are not present in the model or contain gaps or errors, please provide us with a list of corrections and additions you would like to see part of the model.

How can I edit the network visualization and prepare it for publication?
You can edit the node positions by dragging a compound with the mouse. If you want to switch off the effect of the force function on a node you can double click on it. Subsequently you should export the network in SVG format, after which you can further edit it with vector image editors such as Inkscape.

I am not able to download the network in SVG format.
Currently the query results can only be downloaded using the Firefox or Chrome web browser.

The {{browser_name}} browser is not displayed correctly on my mobile phone or tablet.
The performance of the {{browser_name}} browser has not been extensively tested on mobile devices. Nevertheless, it has been shown to work correctly on mobile devices of the latest generation with a minimal screen width of 640 pixels and using the Firefox web browser.

Contact

The {{browser_name}} model and pathway browser have been developed by Jan Bert van Klinken at the Leiden and Amsterdam University Medical Centra. For questions, comments or to report bugs in the browser or errors in the {{browser_name}} model, please send an email to j.b.van_klinken@lumc.nl or j.b.vanklinken@amsterdamumc.nl

Acknowledgements

The {{browser_name}} model and pathway browser have been developed with funding from the Dutch Heart Foundation (CVON2014-02 ENERGISE).