Visualization methods of hierarchical biological data: A survey and review

© 2017 International Ambient Media Association (iAMEA). All rights reserved. The sheer amount of high dimensional biomedical data requires machine learning, and advanced data visualization techniques to make the data understandable for human experts. Most biomedical data today is in arbitrary high dimensional spaces, and is not directly accessible to the human expert for a visual and interactive analysis process. To cope with this challenge, the application of machine learning and knowledge extraction methods is indispensable throughout the entire data analysis workflow. Nevertheless, human experts need to understand and interpret the data and experimental results. Appropriate understanding is typically supported by visualizing the results adequately, which is not a simple task. Consequently, data visualization is one of the most crucial steps in conveying biomedical results. It can and should be considered as a critical part of the analysis pipeline. Still as of today, 2D representations dominate, and human perception is limited to this lower dimension to understand the data. This makes the visualization of the results in an understandable and comprehensive manner a grand challenge. This paper reviews the current state of visualization methods in a biomedical context. It focuses on hierarchical biological data as a source for visualization, and gives a comprehensive survey of visualization techniques for this particular type of data