The library renders both rectangular and circular phylograms from Newick and phyloXML [11] formats. Newick files (outputted by many popular phylogenetic packages, such as PHYLIP [1]) include at most only labels, edge lengths and bootstrap values; the tree renderings are therefore limited to these elements. Implementing the recently devised XML format specialized for phylogentic trees, phyloXML, offers the ability to include features such as text nodes that link to other websites or reveal descriptions on rollovers (see Figure 1). These core characteristics are present in both tree formats.

The javascript library consumes and interprets the raw Newick or phyloXML formats directly, from which trees are then generated. This approach is advantageous because it allows the readership to have access to the underlying tree source. Many online access journals and databases (PLoS [12], BioMed Central [13], Pubmed [14] and others) provide full papers in XML format while the phylogenetic trees themselves are provided as static images. It is of our opinion that the readership of online scientific literature would benefit greatly if the trees were also maintained in XML format. This would both provide greater transparency and allow the trees to be indexed by data-mining tools.

All of the stylistic qualities of the tree may be specified either programmatically as parameters or within the tree's XML file. Characteristics such as the font size or type, branch stroke width, mouse rollover events and scale can be easily customized. This offers the ability for each tree to be uniquely rendered using its predefined style in the XML file, or conversely allows for sites to override styles to maintain trees that are visually similar. Detailed explanations and documentation of available parameters and file formatting are available on our website.

The library integrates charting analytics within the tree visualization. Two charting formats are currently supported: binary and bar charts (see Figure 2). These reporting features are based on provided data within the tree XML file. The software normalizes the data and renders the charts according to user-defined styles in the XML file or parameters set at tree instantiation. jsPhyloSVG is flexible enough to an unlimited number of chart tracks to be added to the tree, with any number of stylistic characteristics.

The limitations of the library are primarily related to the authors' choice of using SVG to render the visualizations. Firstly, the format is not natively supported by current versions of Internet Explorer. This restriction has been overcome by automatically rendering the tree in the analogous format VML for these browsers. Secondly, SVG exposes its elements to the browser's Document Object Model (DOM) in order to allow for interactivity at an added slight computational cost. These computational tasks begin to significantly affect the performance of the browser, however, when rendering large trees that contain more than 2000 nodes (see Table S1). The choice to use SVG was deliberate in order to fulfill the primary objective of providing interactive capabilities despite the cost in performance for large trees.

jsPhyloSVG is a comprehensive open-source solution for rendering dynamic phylogenetic trees. It leverages modern approaches by supporting HTML5 for the widest possible audience. It consumes phylogenetic inferences in Newick or XML formats and generates sophisticated trees that incorporate charting analytics in scalable vector images. The trees are also dynamic and provide instantaneous user feedback as they are navigated. Finally, the library is simple to extend and may be easily integrated into existing web applications.

Project home page: http://www.jsphylosvg.com

Operating System: Platform independent

Programming Language: Javascript

License: Open Source, GPL.