GeneNet VR: Interactive visualization of large-scale biological networks using a standalone headset

Authors

  •  Alvaro Martinez Fernandez
  • Lars Ailo Bongo
  • Edvard Pedersen

Abstract

Visualizations are an essential part of biomedical analysis result interpretation. Often, interactive networks are used to visualize and interpret the data. However, the high interconnectivity, and high dimensionality of the data often results in information overload, making it hard to interpret the results. To address the information overload problem, existing solutions typically either use data reduction, reduced interactivity, or expensive hardware. We propose using the affordable Oculus Quest Virtual Reality (VR) standalone headset for interactive visualization of large-scale biological networks. We present the design and implementation of our solution, GeneNet VR, and we evaluate its scalability and usability using large gene-to-gene interaction networks. We achieve the 72 FPS required by the Oculus' performance guidelines for the largest of our networks (2693 nodes) using both a GPU and the Oculus Quest standalone. We found from our interviews with biomedical researchers that GeneNet VR is innovative, interesting, and easy to use for novice VR users. We believe a?ordable hardware like the Oculus Quest has a big potential for biological data analysis. However, additional work is required to evaluate its bene?ts to improve knowledge discovery for real data analysis use cases. GeneNet VR is open-sourced: https://github.com/kolibrid/GeneNet-VR. A video demonstrating GeneNet VR used to explore large biological networks: https://youtu.be/N4QDZiZqVNY.

Published

2021-11-14