Freia: Exploring Biological Pathways Using Unity3D

Abstract

To understand the biological processes related to the development of cancer there is a need for interactive data visualization tools that integrates experimental data with biological knowledge. Existing visualization tools have shown their usefulness, but next-generation biological data analysis requires both integrating different data types and larger datasets. This requires data exploration tools with a flexible data model that is efficiently represented and processed. Entity Component System in game engines fulfil both requirements. There are two main challenges for using game engines for biological data visualization. First, how to represent the biological data to be visualized using the game engine data structures. Second, how to efficiently implement the necessary data exploration operations using these data structures. This thesis presents the, to our knowledge, first approach for mapping biological data to the ecs model. We used the approach to implement Freia, an application for visualizing gene expression data integrated with pathway images. We evaluated the performance and scalability of Freia by measuring the smoothness of key data exploration operations. Our results show that Freia provides a frame rate above 30 fps for these operations for up to 100 simultaneously shown pathways. We believe our approach demonstrates that game engines are well suited to implement data visualization tools for the upcoming biological data studies.