Dynamic visualisation of million-tip trees: The OneZoom project

The complete tree of life is now available, but methods to visualise it are still needed to meet needs in research, teaching and science communication. Dynamic visualisation of million-tip trees requires many challenges in data synthesis, data handling and computer graphics to be overcome. Our approach is to automate data processing, synthesise data from a wide range of available sources, then to feed these data to a client-side visualisation engine in parts. We develop a way to store the whole tree topology locally in a highly compressed form, then dynamically populate metadata such as text and images as the user explores. The result is a seamless and smooth way to explore the complete tree of life, including images and metadata, even on relatively old mobile devices. The underlying methods developed have applications that transcend tree of life visualisation. For the whole complete tree, we describe automated ID mappings between well known resources without resorting to taxonomic name resolution, automated methods to collate sets of public domain representative images for higher taxa, and an index to measure public interest of individual species. The visualisation layout and the client user interface are both abstracted components of the codebase enabling other zoomable tree layouts to be swapped in, and supporting multiple applications including exhibition kiosks and digital art. After 10 years of work, our tree of life explorer is now broadly complete, it has attracted nearly 1.5 million online users, and is backed by a novel long-term sustainability plan. We conclude our description of the OneZoom project by suggesting the next challenges that need to be solved in this field: extinct species and guided tours around the tree.

Automated summary

The text discusses the availability of the complete tree of life and the need for methods to visualize it for research and teaching. They have developed an effective visualization engine through automated data processing and dynamic population of metadata. Additionally, it mentions some achievements of the project and future challenges to be addressed.