A divide-and-conquer method for scalable phylogenetic network inference from multilocus data.

Area of Science:

• Evolutionary biology
• Computational biology
• Genomics

Background:

• Reticulate evolutionary histories, including hybridization, are best modeled using phylogenetic networks.
• Current statistical inference methods for phylogenetic networks are limited to small numbers of genetic loci and genomes.
• Incomplete lineage sorting is a key evolutionary process that needs to be accounted for in network inference.

Purpose of the Study:

• To introduce a novel, scalable two-step method for inferring phylogenetic networks.
• To enable accurate phylogenetic network inference from multiple unlinked genetic loci.
• To overcome the scalability limitations of existing phylogenetic network inference methods.

Main Methods:

• A two-step approach that infers networks on subproblems and merges them.
• Formulation of a Hitting Set problem to reduce the number of inferred trinets.
• Implementation of a heuristic algorithm to solve the Hitting Set problem.
• Performance evaluation using simulated and biological datasets.

Main Results:

• The novel two-step method accurately infers phylogenetic networks at a large scale.
• The method demonstrates improved accuracy and computational efficiency compared to existing approaches.
• Successful application to both simulated and real biological data.

Conclusions:

• The developed method represents a significant advancement in large-scale phylogenetic network inference.
• This approach facilitates more accurate reconstruction of complex evolutionary histories.
• The findings pave the way for broader applications of phylogenetic networks in evolutionary studies.