Related Experiment Videos
Inferring complex phylogenies using parsimony: an empirical approach using three large DNA data sets for angiosperms.
D E Soltis1, P S Soltis, M E Mort
1Department of Botany, Washington State University, Pullman, Washington 99164-4238, USA. dsoltis@mail.wsu.edu
Systematic Biology
|June 18, 2002
Summary
Combining DNA data sets significantly improves phylogenetic analysis efficiency and accuracy for large plant datasets. Parsimony analysis of combined sequences allows for thorough tree searches and higher clade support, overcoming computational challenges.
Area of Science:
- Phylogenetics
- Molecular Evolution
- Computational Biology
Background:
- Parsimony analysis is a common method for inferring evolutionary trees.
- Analyzing large DNA datasets presents computational and analytical challenges.
- Previous studies have explored combining data but feasibility for very large matrices was unclear.
Purpose of the Study:
- To assess the feasibility of parsimony analysis for large, combined DNA sequence datasets.
- To evaluate the impact of data combination on computational efficiency and phylogenetic accuracy.
- To determine if combining datasets improves signal for phylogenetic inference.
Main Methods:
- Heuristic parsimony searches and bootstrap analyses were performed.
- Separate DNA datasets (18S rDNA, rbcL, atpB) from 190 angiosperms were analyzed.
- These datasets were combined into a single, larger matrix (4,733 bp) for analysis.
Main Results:
- Combined data set analyses showed significantly improved computer run times compared to separate or pairwise combined datasets.
- Tree searches using TBR branch swapping completed rapidly for the combined dataset but not for separate datasets.
- The combined dataset yielded higher internal clade support and more clades with bootstrap support >= 50%.
Conclusions:
- Combining DNA sequence data is a viable solution for overcoming computational and analytical challenges with large datasets.
- Sufficient phylogenetic signal for numerous taxa is best achieved by combining multiple data sources.
- Data combination enhances both the efficiency of phylogenetic searches and the reliability of inferred evolutionary relationships.