Related Experiment Videos
Inferring species phylogenies from multiple genes: concatenated sequence tree versus consensus gene tree
Sudhindra R Gadagkar1, Michael S Rosenberg, Sudhir Kumar
1Department of Biology, University of Dayton, Dayton, Ohio 46469-2320, USA. gadagkar@notes.udayton.edu
Summary
Concatenating multiple gene sequences into a super-gene alignment provides more accurate species trees than combining individual gene phylogenies. This concatenation approach demonstrates high accuracy, even with diverse evolutionary patterns, but warrants careful interpretation of support values.
Area of Science:
- Phylogenetics
- Computational Biology
- Evolutionary Biology
Background:
- Phylogenetic trees are crucial for understanding evolutionary relationships.
- Two primary methods exist for inferring species trees from multiple genes: concatenation and consensus.
- The relative accuracy and potential biases of these methods require thorough investigation.
Purpose of the Study:
- To compare the accuracy of the gene concatenation approach versus the consensus method for species tree inference.
- To evaluate these methods under diverse evolutionary parameters and worst-case simulation scenarios.
- To assess the impact on phylogenetic accuracy across different inference algorithms.
Main Methods:
- Computer simulations were conducted using 448 parameter sets, varying evolutionary rates, sequence lengths, base compositions, and transition/transversion biases.
- Worst-case scenario analysis involved selecting the most error-prone replicate dataset for each parameter set.
- Phylogenetic trees were inferred using both randomly selected and worst-case replicates, primarily with the neighbor-joining (NJ) method, and also parsimony and likelihood.
Main Results:
- The concatenation approach consistently yielded more accurate species trees compared to the consensus method, even with genes exhibiting divergent evolutionary patterns.
- High accuracy (>95%) was achieved with concatenation using as few as 10 genes.
- Concatenation can sometimes reinforce systematic biases, leading to inflated bootstrap support for incorrect clades.
Conclusions:
- Gene concatenation is a more accurate method for inferring species phylogenies than combining individual gene trees.
- While robust, the concatenation approach necessitates caution regarding bootstrap support values.
- Reporting the number of genes supporting a clade in concatenated trees, alongside bootstrap values, is recommended for accurate interpretation.