The Effect of Taxonomic Sampling on Accuracy of Phylogeny Estimation: Test Case of a Known Phylogeny
Year: 1998
Type: article
Abstract: Because of extinction, undiscovered species, and constraints of time, money and material, most molecular phylogenetic analyses are performed without extensive coverage of the group of interest.The degree of taxonomic coverage can affect hypotheses of tree topology (e.g., Gauthier, Kluge, and Rowe 1988), ancestral state assignment (Sanderson 1996), character evolution (Donoghue et al. 1989), and character correlation (Sille ´n-Tullberg 1993).Studies are beginning to examine the effect of taxon sampling on accuracy (e.g., Huelsenbeck 1991;Wheeler 1992;Wiens and Reeder 1995;Kim 1996).The favorability of including more taxa is now a working assumption for many (e.g., Baverstock and Moritz 1996, p. 26).This paper will examine the relationship of taxon sampling to accuracy using the known phylogeny of T7 bacteriophage of Hillis et al. (1992).Hillis et al. (1992) created a known phylogeny of the T7 bacteriophage by serially propagating strains in the presence of a mutagen.For this first experimental phylogeny they enforced completely symmetrical branching and equal time between splitting events.This system has proven useful for questions of performance of phylogenetic methods (Hillis et al. 1992;Crandall 1994; Hillis and Huelsenbeck 1994;Wiens and Reeder 1995), molecular evolution (Bull et al. 1993;Cunningham et al. 1997), and confidence in phylogeny (Hillis and Bull 1993).Details of this phylogeny are available in Hillis et al. (1992), Bull et al. (1993), and Sober (1993).The advantages of the T7 system are that (unlike in studies of literature data sets) the phylogeny is known and (unlike in simulations) the tree is the result of biological evolution in a lineage of organisms (Hillis 1995).Disadvantages of this system have to do with applicability.For example, molecular evolution in T7 viruses in a controlled experiment may not be a useful model for evolution in many ''higher'' organisms under reallife conditions, and the single tree shape and set of branch lengths may not be representative of most phylogenies (although subsampling taxa and characters can ''create'' different phylogenetic conditions).I perform subsampling experiments of taxa and characters from the T7 restriction site data to examine the effect of taxonomic sampling at three levels: (1) Does sampling more taxa improve or worsen accuracy 1 Present address: Division of Amphibians and Reptiles, National Museum of Natural History, Smithsonian Institute, Washington, D.C.
Source: Molecular Biology and Evolution
Author: S. Poe
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: bronze
Funder: National Science Foundation
OpenAlex ID: W2332869178
