Fresh data plots sprout on Nature’s latest cover

Background. Most deaths from cancer are related to tumours spreading to secondary sites in the body through metastasis, yet there are significant gaps in our knowledge of the underlying biology of this process. In this week’s issue of Nature, Todd Golub and his colleagues report the MetMap, a barcoding system that they have used to determine the metastatic potential of human cancer cell lines. From their analyses, the researchers created petal plots, as illustrated on the cover, that relate to the metastatic pattern of the cancer cells.

Creating a new visualization technique for cancer research. The petal plots caught our attention as a powerful way to show layers of information at a glance. This is what visuals do best. And perhaps naturally, we thought the technique would make an excellent cover subject. Andrew Tang, a designer at the Broad Institute of MIT and Harvard, describes his team’s work developing the plots, and their thinking behind their cover art:

“Petal Plot is a novel visualization originally designed for representing and exploring MetMap data. The petal plot renders data in a form that resembles a flower and the number of petals is determined based on a categorical attribute of data. Each petal can encode two numerical properties as its length and width. The shape of the petal naturally creates visual weight emphasizing the relationship between the two properties, allowing informative trends and patterns to emerge effectively.

Figure 1 from the paper displays data in Petal Plot form. (Source: Nature)

Utilizing the flower as a visual metaphor achieves a familiar, intuitive and memorable image without sacrificing data accuracy or omitting key attributes. Furthermore, due to its compact radial orientation, a large and multivariate data collection, such as MetMap, can be described and visualized as an array of petal plots in a single view. As a result, this new visualization provides a robust and elegant summary, enabling data comparisons and pattern recognition for complex and interconnected data.

After observing the vast variety and number of petal plots from the MetMap data, and thinking of these as flowers, we began to see more ways to expand the potential of this visualization. For instance, additional attributes from the data can be accommodated and rendered into the 3D space, resulting in an image that recalls a flower field.

This led us toward the concept for the cover design of this issue. We were further inspired by contemporary woodblock prints as the geometric shapes mirror the abstract visual encoding of complex scientific data. Stylistically woodblock prints capture a technique that indexically shows organic forms through its process; the finished textures, and the overlaps of printing.

An early pencil sketch of the concept (left) and a computer render of plots on a data field (right) were produced in the cover development process.

The final version of the cover juxtaposes nature with technology, design with engineering, and the meaning of MetMap data with the aesthetics of its representation.”

The final cover adopts a woodblock artistic style, with a minimalist palette of rich colours, blending themes of the natural world, data and technology.

With thanks to the team at the Broad Institute who contributed to the cover: Xin Jin, Mary O’Reilly, Lia Petronio and Andrew Tang.