Reference Genome Choice and Filtering Thresholds Jointly Influence Phylogenomic Analyses.

Area of Science:

• Genomics
• Bioinformatics
• Evolutionary Biology

Background:

• Molecular phylogenies are crucial for understanding diversification, trait evolution, and biogeography.
• Genomic data processing can introduce biases into phylogenetic analyses.
• The influence of reference genome choice and bioinformatic parameter interactions on phylogenetic inference is not well understood.

Purpose of the Study:

• To investigate how reference genome choice affects phylogenetic inference.
• To examine the interaction between reference genome choice, bioinformatic filtering, and phylogenetic methods.
• To assess the impact of different filtering strategies on phylogenetic accuracy.

Main Methods:

• Utilized simulated and empirical genomic datasets.
• Analyzed phylogenetic inference under varying reference genomes and bioinformatic filtering parameters.
• Evaluated tree topology, imbalance, and center of gravity.

Main Results:

• More stringent minor allele filters (MAC) bias inferred trees away from true species topologies.
• Optimal topological accuracy was achieved with MAC >3-4 filters for 51-taxa datasets.
• Filtering for missing data improved topology accuracy but could distort mutation spectra.

Conclusions:

• Reference genome selection is a critical bioinformatic decision with downstream implications for phylogenetic analyses.
• Bioinformatic filtering choices, particularly minor allele filters, significantly influence phylogenetic inference accuracy.
• The optimal approach for assembling and filtering genomic data depends on study system and dataset attributes.