Related Experiment Videos
When is it safe to use an oversimplified substitution model in tree-making?
1Institute of Molecular Evolutionary Genetics, Pennsylvania State University, University Park 16802, USA.
Molecular Biology and Evolution
|November 1, 1996
Summary
The p-distance is a reliable estimator for evolutionary distances and tree-making under a molecular clock. However, it can lead to incorrect evolutionary trees when substitution rates vary across lineages.
Area of Science:
- Computational Biology
- Evolutionary Biology
- Bioinformatics
Background:
- Selecting optimal mathematical models for evolutionary distance computation is challenging for large datasets.
- Investigators often default to simpler models due to software limitations.
Purpose of the Study:
- To evaluate the effectiveness of the p-distance as an estimator of evolutionary distance for phylogenetic tree construction.
- To assess the performance of p-distances under different evolutionary models, particularly the molecular clock hypothesis.
Main Methods:
- Analysis of the p-distance's properties as an estimator for tree-making.
- Investigating p-distance performance under the molecular clock (stationary, time-reversible Markov model).
- Examining p-distance behavior when substitution rates vary among lineages.
Main Results:
- P-distances enable consistent tree-making when sequence evolution follows a molecular clock.
- P-distances are efficient in recovering correct tree topology under a molecular clock.
- P-distances can result in statistically supported but incorrect trees when substitution rates are not constant.
Conclusions:
- The use of oversimplified models like p-distance is inadvisable without preliminary analysis.
- A practical approach for selecting optimal nucleotide substitution models is proposed.
- A prior distribution estimate for Jukes-Cantor model branch lengths is obtained.