Accurate large-scale phylogeny-aware alignment using BAli-Phy.

Area of Science:

• Computational Biology
• Bioinformatics
• Evolutionary Biology

Background:

• Bayesian phylogenetic analysis (BAli-Phy) is a robust method for co-estimating sequence alignments and phylogenetic trees.
• Its computational intensity has historically limited its application to smaller datasets (approx. 100 sequences).

Purpose of the Study:

• To adapt BAli-Phy for large-scale sequence alignment by repurposing it as a phylogeny-aware alignment method.
• To evaluate the accuracy and scalability of this novel approach compared to existing methods.

Main Methods:

• Estimating a phylogeny from unaligned sequences.
• Utilizing the estimated phylogeny as a fixed tree within BAli-Phy for alignment.
• Testing the method on datasets up to 1000 sequences.

Main Results:

• The adapted BAli-Phy method achieves high accuracy in sequence alignment.
• It significantly outperforms Prank, a leading phylogeny-aware alignment method.
• It is also more accurate than MAFFT, a widely used top-performing alignment tool.
• The method demonstrates scalability for large datasets (up to 1000 sequences).

Conclusions:

• Repurposing BAli-Phy as a phylogeny-aware alignment method offers a highly accurate and scalable solution for biological sequence alignment.
• This approach represents a significant advancement over current popular methods like Prank and MAFFT for large datasets.