Reconstructing phylogeny from reduced-representation genome sequencing data without assembly or alignment.

Area of Science:

• Genomics
• Bioinformatics
• Evolutionary Biology

Background:

• Reduced-representation genome sequencing (RADseq) offers cost-effective genomic analysis but presents challenges for traditional phylogenomic methods due to missing data.
• Alignment-free methods are efficient for whole-genome phylogenetics, particularly for non-model organisms, but their application to RADseq data remains unexplored.

Purpose of the Study:

• To evaluate the efficacy of a full-genome, alignment-free method (AAF) for phylogenetic reconstruction using RADseq data.
• To develop and validate read selection procedures to mitigate missing data issues in RADseq datasets.
• To compare the performance of the alignment-free AAF method with alignment-based approaches for RADseq data.

Main Methods:

• Application of the alignment-free AAF method to RADseq data.
• Development of two read selection protocols to filter RADseq reads based on restriction site presence.
• Validation using simulated datasets and two empirical RADseq datasets.

Main Results:

• Read selection significantly improved phylogenetic reconstruction accuracy across all simulations and real datasets.
• The AAF method, with read selection, performed comparably to or better than alignment-based methods.
• The alignment-free approach demonstrated substantially lower computational requirements compared to alignment-based methods.

Conclusions:

• The AAF pipeline, adapted for RADseq (phyloRAD), provides an accurate and computationally efficient solution for phylogenetic analysis of reduced-representation genomic data.
• Read selection is a crucial step for optimizing phylogenetic inference from RADseq data using alignment-free methods.
• This approach enhances the utility of RADseq for population genetics and interspecific phylogeny, especially for non-model organisms.