Visibility graph based temporal community detection with applications in biological time series.

Area of Science:

• Computational Biology
• Network Science
• Data Analysis

Background:

• Biological systems exhibit complex temporal behaviors crucial for understanding their function.
• Representing time series data as networks is a promising approach, but existing methods struggle with biological data characteristics.
• Detecting dynamic patterns (temporal communities) in these network representations remains a challenge.

Purpose of the Study:

• To develop a novel network-based method for analyzing biological time series.
• To create a robust network representation capable of handling unevenly sampled data and capturing key events like peaks and troughs.
• To introduce an effective algorithm for detecting temporal communities within these specialized networks.

Main Methods:

• Introduction of the Weighted Dual-Perspective Visibility Graph (WDPVG) for constructing networks from time series.
• Utilizing shortest path analysis to identify central nodes (hubs) within the network.
• Aggregating peripheral nodes to the nearest central hubs for temporal community assignment.

Main Results:

• The WDPVG method successfully characterizes unevenly sampled biological time series.
• The community detection algorithm effectively identifies dynamic patterns and events within the time series networks.
• Validation through simulations and application to real biological data demonstrates the method's efficacy.

Conclusions:

• The WDPVG offers a powerful new tool for the analysis of temporal behavior in biological systems.
• This network-based approach enhances the detection of dynamic patterns and events in complex biological data.
• The method provides a robust framework for understanding temporal dynamics in biological research.