Multilocus Sequence Analysis and Bacterial Species Phylogeny Estimation
Pablo Vinuesa
from: Molecular Phylogeny of Microorganisms (Edited by: Aharon Oren and R. Thane Papke). Caister Academic Press, U.K. (2010)
Abstract
This chapter presents a review of critical factors that have to be considered and evaluated in multilocus sequence analysis (MLSA) in order to make robust estimates of bacterial species phylogenies. The theoretical arguments in favor of the conditional data combination will be presented. I will briefly review criteria for marker selection, and will provide practical advice on the computational aspects, potential pitfalls, and software choices available for each step in a MLSA. For this purpose, a detailed case study using atpD, glnII, recA and rpoB sequences of symbiotic root nodule bacteria of the genus Bradyrhizobium will be presented. I will discuss and illustrate the use of phylogenetic congruence analysis, and a strategy to evaluate the additivity of the phylogenetic signals in the different partitions. The importance of using multiple isolates per species/lineage, proper model selection and thorough tree searches to get a good estimate of a multispecies phylogeny will be emphasized. Maximum likelihood and Bayesian phylogeny estimation using supermatrices are thoroughly discussed and critically compared with the new, concatenation-independent, Bayesian Estimation of Species Trees (BEST) algorithm, which is based on the multispecies coalescent read more ...
