PVTree: A Sequential Pattern Mining Method for Alignment Independent Phylogeny Reconstruction.

Area of Science:

• Bioinformatics
• Computational Biology
• Evolutionary Biology

Background:

• Phylogenetic trees are crucial for understanding evolutionary relationships.
• Traditional multiple sequence alignment methods are computationally intensive and require extensive parameter tuning.
• Existing alignment-free methods often overlook global sequence similarities and valuable features like frequent patterns.

Purpose of the Study:

• To develop a novel, accurate, and efficient alignment-free algorithm for phylogenetic tree construction.
• To address limitations of current alignment-free methods, particularly regarding global similarities and frequent patterns.
• To improve phylogenetic inference for large and highly divergent sequence datasets.

Main Methods:

• An alignment-free algorithm based on sequential pattern mining is proposed.
• Sequences are converted into binary representations of sequential patterns.
• Phylogenetic trees are constructed by clustering a distance matrix derived from pattern vectors.
• Pattern weighting and redundancy filtering are incorporated to enhance accuracy for divergent sequences.

Main Results:

• The proposed method demonstrates superior performance compared to other alignment-free approaches.
• Outperformance is particularly notable for large sequence datasets with low similarity.
• Simulated and real-world data validate the effectiveness of the algorithm.

Conclusions:

• The sequential pattern mining-based algorithm offers a robust and accurate alternative for phylogenetic tree construction.
• The method effectively captures global sequence similarities and frequent patterns, improving evolutionary inference.
• This approach provides a significant advancement for analyzing large and divergent genomic datasets.