PathWhiz is a comprehensive, online pathway drawing tool for generating biochemical and biological pathways. It uses publicly accessible databases and easily expandable palettes consisting of pre-drawn pathway components. This protocol describes how to easily build new pathways, replicate and edit existing pathways, and propagate previously drawn pathways to different organisms.
PathWhiz is a web server built to facilitate the creation of colorful, interactive, visually pleasing pathway diagrams that are rich in biological information. The pathways generated by this online application are machine-readable and fully compatible with essentially all web-browsers and computer operating systems. It uses a specially developed, web-enabled pathway drawing interface that permits the selection and placement of different combinations of pre-drawn biological or biochemical entities to depict reactions, interactions, transport processes and binding events. This palette of entities consists of chemical compounds, proteins, nucleic acids, cellular membranes, subcellular structures, tissues, and organs. All of the visual elements in it can be interactively adjusted and customized. Furthermore, because this tool is a web server, all pathways and pathway elements are publicly accessible. This kind of pathway "crowd sourcing" means that PathWhiz already contains a large and rapidly growing collection of previously drawn pathways and pathway elements. Here we describe a protocol for the quick and easy creation of new pathways and the alteration of existing pathways. To further facilitate pathway editing and creation, the tool contains replication and propagation functions. The replication function allows existing pathways to be used as templates to create or edit new pathways. The propagation function allows one to take an existing pathway and automatically propagate it across different species. Pathways created with this tool can be "re-styled" into different formats (KEGG-like or text-book like), colored with different backgrounds, exported to BioPAX, SBGN-ML, SBML, or PWML data exchange formats, and downloaded as PNG or SVG images. The pathways can easily be incorporated into online databases, integrated into presentations, posters or publications, or used exclusively for online visualization and exploration. This protocol has been successfully applied to generate over 2,000 pathway diagrams, which are now found in many online databases including HMDB, DrugBank, SMPDB, and ECMDB.
Biological pathway diagrams are like blueprints for life scientists. They are perhaps the most concise and informative routes for depicting biological processes and the contextual connections between genes, proteins, and metabolites. This is because images are much more efficiently processed and often much more easily understood by humans than text1. The quality, detail, and content of pathway diagrams can vary considerably. These differences often depend on the intended purpose of the pathway and the skills of the pathway artist. Pathways created for educational purposes such as wall charts or textbooks are often created by professional artists. As a result, these pathway diagrams are much more visually pleasing and offer considerably more biological detail with full depictions of metabolite structures, subcellular components, cell structures, tissues, and organs. These "textbook" representations often include detailed notes and commentaries. On the other hand, pathway diagrams designed for internet applications often have to sacrifice artistry and visual richness in favor of simplified machine readable "wiring" diagrams. These wireframe diagrams are more easily image-mapped and hyperlinked. Simplified pathway diagrams are the basis to such popular online pathway databases as KEGG2, MetaCyc3, Wikipathways4, and Reactome5. The emergence of computer-compatible pathway databases has also led to the appearance of computer-compatible pathway drawing tools. In other words, one does not have to be a professional artist or a professional programmer to generate useable pathway diagrams. For instance, BioCyc's Pathway tools6 and Wikipathway's PathVisio software7allow users to freely generate and share machine-readable pathways in BioPAX8 and/or HTML format. Additionally, there are a number of other stand-alone freeware packages as well as commercial packages that support the generation of various machine-readable, wire-frame pathways, such as Cytoscape9, GenMAPP10, PathCase11, and VisANT12.
The simplification of internet pathway diagrams largely grew from the historical limitations found in many web browsers and web-based rendering tools. However, significant advances in web technologies have been made in the past few years. This suggested that it could be possible to generate interactive, internet-compatible pathway diagrams that are just as colorful, just as aesthetically pleasing and just as biologically complete as those found in textbooks. This work led to the development of PathWhiz. PathWhiz was implemented using a Ruby on Rails (http://rubyonrails.org, version 4.2.0) web framework incorporating a MySQL relational database (https://www.mysql.com, version 5.1.50) to manage all of the pathway data, including entity relationships, external references, descriptions, visualization specifications and chemical structures. The front-end web client is controlled by Ruby on Rails combined with Backbone.js (http://backbonejs.org, version 1.0.0) as the front-end web framework for the editor.
Originally developed for the curation of the human-only Small Molecule Pathway Database (SMPDB)13, PathWhiz14has since been extended to support pathway generation for many other organisms and to function as a general pathway image and knowledge database. In particular, this web tool allows for the creation of the full range of biochemical/biological pathways including metabolic, protein interaction, molecular signaling, physiological, and drug/disease pathways. This pathway drawing tool differs from most other pathway generating tools in three major ways: 1) it is a web server rather than a stand-alone, installable software package; 2) it supports the facile generation and interactive visualization of chemical compounds, proteins, nucleic acids, cellular membranes, subcellular structures, tissues, and organs; and 3) it allows users to easily borrow, build or improve upon the work of other users, thereby allowing "crowd-sourced" pathway generation. As a web server, it has several advantages over downloadable, platform-specific software tools. In particular, it is compatible with any platform, operating system and modern web browser. Furthermore, it does not require the user to register in order to begin creating a pathway (though users can freely create a "private account" in order to track and control the accessibility of the pathways they create). Perhaps the most appealing feature of this tool is the amount of biological and biochemical detail that can be easily added into any pathway through a palette of pre-rendered images and an extensive database of protein and biochemical data. This allows both "non-artists" and "non-programmers" to easily create colorful, aesthetically pleasing and richly detailed pathways that are web-compatible and fully machine-readable. A more detailed comparison between PathWhiz and other pathway drawing tools is provided in Table 1.
A number of popular life science databases have already used this pathway drawing tool to create database-specific, online, interactive pathway diagrams. For instance, the Escherichia coli Metabolome Database (ECMDB)15recently updated its pathway library with more than 1,650 pathways drawn using web based tool. Each pathway in the ECMDB is now displayed as a richly colored, fully hyperlinked image map with detailed metabolite and protein structure depictions, as well as a simplified black and white KEGG-like wire diagram. This large-scale pathway update led to the discovery of many intermediate metabolites that had not been previously included in other Escherichia coli metabolic databases. Other databases, such as the Human Metabolome Database (HMDB)15, not only rely on PathWhiz pathways to depict and describe metabolic and signaling pathways, but also to depict the metabolic changes involved in diseases such as cancer. HMDB currently includes 101 metabolic pathways, 376 drug action pathways, 233 disease-associated pathways and 16 signaling pathways, all generated through this web tool.
The following protocol describes in detail how PathWhiz can be used to easily create, replicate, and propagate biochemical pathways for a variety of purposes and applications.
The protocol described here for creating a simple metabolic pathway (the TCA cycle) can be adapted to create a wide variety of machine-readable, biologically complex pathways for any species. Furthermore, this protocol also describes how one can replicate or propagate existing pathways created by other users. Constructing a pathway using this tool requires repeated step-by-step additions of reactions, interactions, transport processes, and sub-pathways, each of which are connected by overlapping elements. Putting all of these together allows one to create colorful, visually pleasing pathway diagrams that provide considerable biological detail and useful biological context. The steps described in this protocol are relatively simple, and the time it takes to build a pathway diagram depends on the size and complexity of the pathway. With a bit of practice, most individuals can render a high quality pathway diagram consisting of about 15-20 reactions or processes and several cellular components in about 15 min. A brand new user may take up to 30-40 min to generate a pathway of similar size and complexity. The time needed to generate a pathway is roughly linear with the number of reactions/processes that need to be rendered.
Creating a high quality pathway through this web-based tool depends on the quality and detail of the source material (pathways from books, online databases, experimental data, hand-drawn sketches) and the fastidiousness of the pathway "artist". Those wishing to generate higher quality pathway diagrams should pay particular attention to sections 1.11, 1.15 and 1.20 of the protocol, since these sections describe the creation and editing of reaction elements (reactants/products, enzymes, edges, images, zoom boxes, labels, and membranes). The best pathway diagrams will intelligently amalgamate information from as many existing representations of the pathway as possible including those found in books, posters, papers, and online databases. Another key to generating high quality pathways is carefully checking the correctness of the reactions before creating a reaction (through section 1.11 of the protocol). Taking the time and effort to ensure the reactants, products, and enzymes involved (many of which already exist in PathWhiz's large database) are correct for each species is very important. It is also important to be aware of the cellular location of the reactions and to include key cellular or subcellular components in order to provide the correct biological context. This can be done by checking and corroborating the reaction through online databases such as UniProt. Having all the required information at hand, along with a rough, hand-drawn sketch of the pathway to be created will greatly reduce errors and the overall time spent on drawing or rendering.
As might be expected, larger and more complex pathways will take longer to render, particularly if the desired elements and processes are not already in PathWhiz's database. When working with larger pathways, it is usually wise to switch from the Autosave mode to the Manual save mode, in order to prevent a long lag between actions. When replicating a pathway, the amount of time the user might wait for the pathway to be generated depends on the number of elements in the pathway. Most pathways can be replicated in about 1-2 min. When propagating a pathway, successful rendering of the newly propagated pathway depends on how similar the two species are, as PathWhiz uses BLAST17 sequence searches to find homologous enzymes between species. Larger pathways will be slower to propagate because BLAST will need to be run on a greater number of enzymes. Attempting to propagate pathways between significantly dissimilar species (say between yeast and humans) will result in pathways being rendered with a number of unknown proteins. These "distantly" propagated pathways will usually require additional manual editing. Due to the highly visual nature of pathway diagrams and the detail that can be brought to a pathway, it is always a good idea to work on a computer with a reasonably large screen (>20 inches or >50 cm) and a good internet connection (>5 mbps).
If problems with rendering or screen refreshing are encountered, the user may have to do a small amount of troubleshooting. If a large, complex pathway takes too long to update, the user may have to refresh the page. If a pathway does not propagate as expected, the user may have to do some manual editing to ensure that all elements are displayed correctly. Also, as a more specific example, if elements of a reaction are not displayed, the user may have to make sure that all elements or enzymes are properly selected and the edges are not hidden. The "Help" link on the main header may be useful if a problem is encountered. A tutorial is available under the "Tutorial" tab and a user manual is available under "User Guides" tab. Both explain many of the tool's features in detail. The User Guide may be used to troubleshoot or explain potential limitations for a particular feature, such as when a user locks a pathway and later wishes to edit it.
As highlighted through this protocol and through the examples provided in the accompanying figures, this tool offers a number of unique features not found in any (or most) other pathway drawing tools (see Table 1). First, it is fully web-based and completely platform-independent. Second, it supports the rendering and facile generation of multicolored, biologically complex, visually pleasing, fully hyperlinked pathway diagrams that can also be converted to machine-readable formats (BioPAX, SBGN-ML18, SBML19, PWML14). Third, pathway diagrams generated by this tool can be browsed, searched, selected and easily explored through an easy-to-use online database and viewing interface. Fourth, the web tool is designed to support community pathway contributions, allowing for "pathway crowd sourcing" that encourages the sharing and generation of new pathways and new pathway elements.
Pathways generated by this web based tool can be used for a variety of applications. Richly detailed, fully hyperlinked pathways can be readily integrated into organism-specific databases for proteomics, metabolomics or systems biology applications. Internet-accessible pathways are especially useful for education and training purposes, as the details available through web-based images are often much greater than what can be displayed via a static image or through a single textbook or journal page. This web based tool also supports the generation of pathway representations that are more suited for printing and publication. As a result, many images generated by this web based tool are appearing in papers, posters and slide presentations. Exporting pathways into text-based data exchange file formats (such as BioPAX and SBML) allows for pathways generated using this web server to be used directly in computational analysis for systems biology or metabolic modeling applications. Propagating pathways between species allows inferences to be made about biological processes, especially among those species that have been very recently sequenced. While not all existing pathways currently exist in PathWhiz, its public pathway database continues to grow, leading to the emergence of new, crowd-sourced pathway collections. These collections will not only be readily extendable to new species, they will hopefully lead to a deeper understanding of their unique biology and biochemistry.
The authors have nothing to disclose.
The authors wish to thank the Canadian Institutes of Health Research (CIHR) and Genome Alberta, a division of Genome Canada, for financial support.
Computer with colour screen | N/A | N/A | >20 inches or >50 cm |
Internet connection | N/A | N/A | >5 mbps |
Modern web browser | N/A | N/A | Google Chrome (v. 31 and above), Internet Explorer (v. 9 and above), Safari (v. 7 and above), Opera (v. 15 and above) and Firefox (v. 23 and above) |