Developing new algorithms and software for explaining complex structured data in biology: sonification and visualisation of phylogenetic trees

Abstract:

This interdisciplinary project will develop algorithms and software combining sonification and visualisation, to help students better understand complex, tree-structured data and processes. It will use a case study in life sciences, of cyanobacterial phylogeny and ancestral state reconstruction. Background:

A phylogenetic tree illustrates relations among different species [1]. However, most software can only produce large static images of phylogenies [2]. For thousands of species (as is now possible with current technology), images are complicated and difficult for non-expert audiences to understand.

In recent years, tools for 3D visualisation (3D visual representation of data) and sonification (audio representation of data) have shown promise in various applications, assisting understanding of complex data in education and public engagement [3,4]. However, there is relatively little research on the effectiveness of the combined use of visualisation and sonification.

We propose that new algorithms and approaches, such as allowing the user to visually highlight part of the tree using sound to convey local properties, will present a step-change in exploration and understanding of phylogenetic data; and that findings may be generalised to communication of complex tree-structured data in general.

Cyanobacteria are a diverse, cosmopolitan group of bacteria [5]. Combining evolutionary biology with software engineering, we propose to investigate ancestral states of cyanobacterial traits and habitat as a case study to develop a generally applicable phylogenetic tree sonification and visualisation software. We will develop and evaluate this for a target audience of secondary school students from an educational perspective.

Phylogeny is on the curriculum in Scotland (Higher Biology) [6], however, based on communications with secondary school teachers, they find it difficult to teach at this level. The software and methodology are expected to have wide application, also in higher education and research, within and beyond life sciences applications.

Key research questions:

1. Can we represent a cyanobacterial phylogenetic tree (and associated trait data) dynamically via 3D visualisation and sonification tools? 2. Does this representation assist in students' understanding of phylogeny and trait evolution?

3. How, if at all, does learning through the dynamic phylogenetic tree method (3D picture and sound) differ compared to with more traditional approaches and materials?

4. Can the visualisation and sonification methods and algorithms be implemented in a general software for use with any species, not just cyanobacteria? 5. Can the tool be successfully applied to tree-structured data in general, beyond the life sciences application? Methodology:

The project combines phylogenetic analysis, visualisation and sonification tool (software) development and a combination of qualitative data collection methods including focus groups, semi-structured interviews, and classroom observations. Based on the existing phylogenetic tree building tools and libraries, we will develop an algorithm to visualise and sonify the traditional phylogenetic tree using programming languages (i.e.

Python and Java). Evaluation of the effectiveness of the approach for secondary school students will be carried out by qualitative methods. In addition, before- and after- multiple choice questions will be asked to see the changes of students' ideas of evolution and phylogeny due to the activities will be subject to quantitative analysis with applying statistic models in R.