Measuring community similarity with phylogenetic networks

Area of Science:

• Ecology and Evolutionary Biology
• Bioinformatics and Computational Biology

Background:

• Assessing community similarity traditionally relies on species composition, with recent advancements incorporating phylogenetic information.
• Phylogenetic relationships are crucial for understanding evolutionary divergence and its impact on community structure.
• Existing phylogenetic measures may not fully capture complex evolutionary histories, especially when incongruence is present.

Purpose of the Study:

• To extend existing β diversity measures (e.g., Bray-Curtis, UniFrac) to be computable on phylogenetic networks.
• To develop and present a novel methodology for calculating community similarity that accounts for phylogenetic uncertainty and conflict.
• To introduce freely available software, Network Diversity, for implementing these network-based diversity measures.

Main Methods:

• Developed network-based measures for β diversity, focusing on phylogenetic split systems derived from median network and neighbor-net methods.
• Implemented 11 qualitative and 14 quantitative network-based community similarity measures within the Network Diversity software.
• Applied network-based diversity approaches to diverse datasets: pneumococcal isolates, human mitochondrial DNA, and proteorhodopsin sequences.

Main Results:

• Network-based measures successfully recovered major expected patterns of variation across all tested datasets.
• Results demonstrate the robustness of community patterns to phylogenetic uncertainty and conflict when using network approaches.
• Network-based measures can differ significantly from tree-based or non-phylogenetic methods when phylogenetic signals are incongruent.

Conclusions:

• Network-based diversity measures provide a robust methodology for assessing community similarity in the presence of phylogenetic uncertainty.
• These methods allow for the calculation of β diversity over complex phylogenetic network structures, like median networks.
• The Network Diversity software facilitates the application of these advanced ecological analyses to diverse biological datasets.