A phylogenetic analysis of the brassicales clade based on an alignment-free sequence comparison method

Area of Science:

• Computational Biology
• Phylogenetics
• Genomics

Background:

• Phylogenetic trees are crucial for understanding species evolution, typically inferred from sequence alignments.
• Aligning whole eukaryotic genomes and computing phylogenetic trees are computationally intensive challenges.
• Existing methods face limitations in speed and scalability for large genomic datasets.

Purpose of the Study:

• To develop and apply an alignment-free method for phylogenetic tree inference in the Brassicales clade.
• To overcome computational limitations associated with traditional alignment-based phylogenetic analyses.
• To assess the utility of Frequency Chaos Game Representation (FCGR) for phylogenetic classification.

Main Methods:

• Utilized Chaos Game Representation (CGR) to generate unique sequence fingerprints.
• Employed Frequency Chaos Game Representation (FCGR) by dividing CGRs into grid squares representing oligonucleotide frequencies.
• Inferred phylogenetic trees using distance measures between FCGRs from whole genomes, EST data, and mitochondrial genomes of Brassicales species.
• Applied Euclidean distance, Fitch-Margoliash, and Neighbor Joining algorithms for tree construction.
• Introduced bootstrap re-sampling for assessing the support of phylogenetic branchings based on FCGRs.

Main Results:

• Phylogenetic trees reconstructed using FCGRs and Euclidean distance showed general agreement with single-gene trees.
• Fitch-Margoliash and Neighbor Joining algorithms produced similar or identical phylogenetic trees.
• Bootstrap re-sampling was successfully applied to FCGR-based trees, providing branch support information.
• FCGRs proved to be fast to calculate and effective in phylogenetic inference.

Conclusions:

• The alignment-free FCGR method provides an efficient alternative for phylogenetic tree construction in the Brassicales clade.
• FCGR-based phylogenetic trees are comparable to those derived from traditional methods and single-gene analyses.
• FCGRs can supplement alignment-based data and morphological characteristics to enhance phylogenetic classification, especially in ambiguous cases.