Performance comparison between k-tuple distance and four model-based distances in phylogenetic tree reconstruction.

Area of Science:

• Bioinformatics
• Computational Biology
• Evolutionary Biology

Background:

• Phylogenetic tree reconstruction relies on multiple sequence alignment (MSA), which is computationally intensive and prone to errors.
• Errors in MSA can compromise the accuracy of subsequent phylogenetic tree construction.

Purpose of the Study:

• To evaluate the efficacy of k-tuple distance for phylogenetic tree reconstruction as an alternative to traditional MSA-based methods.
• To compare the accuracy and efficiency of k-tuple distance against established model-based distance estimators.

Main Methods:

• Extended the application of k-tuple distance for direct phylogenetic tree reconstruction.
• Compared k-tuple distance with Jukes-Cantor, Kimura, F84, and Tamura-Nei estimators using 1470 simulated sequence sets.
• Constructed neighbor-joining and BioNJ trees to assess performance.

Main Results:

• Trees built using k-tuple distance demonstrated higher accuracy compared to other methods in most scenarios.
• Accuracy gains were particularly significant (up to twofold or more) for highly divergent sequences.
• K-tuple distance obviated the need for MSA, leading to substantial time savings.

Conclusions:

• K-tuple distance is a robust and efficient method for phylogenetic tree reconstruction, outperforming traditional model-based estimators.
• This approach mitigates MSA-related errors and accelerates phylogenetic analysis, especially for large datasets.