Related Experiment Videos
Detecting interspecific recombination with a pruned probabilistic divergence measure
Dirk Husmeier1, Frank Wright, Iain Milne
1Biomathematics and Statistics Scotland, JCMB The King's Buildings, Edinburgh, UK. dirk@bioss.ac.uk
Bioinformatics (Oxford, England)
|December 2, 2004
Summary
This study introduces a pruning method to improve interspecific recombination detection in DNA sequences. The new approach enhances accuracy by addressing issues with diffuse tree topology distributions in sliding-window analyses.
Area of Science:
- Bioinformatics
- Computational Biology
- Molecular Evolution
Background:
- Sliding-window methods for detecting interspecific recombination rely on monitoring changes in tree topology distributions.
- Increased taxa or decreased window size can cause distribution diffusion, reducing detection accuracy.
Purpose of the Study:
- To investigate a post-processing pruning method to enhance recombination detection accuracy.
- To address the limitations of diffuse posterior distributions in sliding-window analyses.
Main Methods:
- A pruning method based on post-processing clustering was developed.
- The Robinson-Foulds distance was used as a metric in tree topology space.
- The method was applied to synthetic and real-world DNA sequence alignments.
Main Results:
- The pruning method demonstrated significant improvement in recombination detection.
- Performance was compared against established methods like Recpars and DSS.
- The effectiveness of the pruning strategy was validated on diverse datasets.
Conclusions:
- The proposed pruning method effectively redeems the shortcomings of standard sliding-window approaches.
- This technique offers a more accurate and robust solution for detecting interspecific recombination.
- The study provides a valuable advancement in phylogenetic analysis tools.