Publications [#184093] of Allen G. Rodrigosearch PubMed.
- Ross, Howard A. and Nickle, David C. and Liu, Yi and Heath, Laura and Jensen, Mark A. and Rodrigo, Allen G. and Mullins, James I., Sources of variation in ancestral sequence reconstruction for HIV-1 envelope genes.,
Evolutionary bioinformatics online, vol. 2
pp. 53--76  .
(last updated on 2010/12/13)
We characterized the variation in the reconstructed ancestor of 118 HIV-1 envelope gene sequences arising from the methods used for (a) estimating and (b) rooting the phylogenetic tree, and (c) reconstructing the ancestor on that tree, from (d) the sequence format, and from (e) the number of input sequences. The method of rooting the tree was responsible for most of the sequence variation both among the reconstructed ancestral sequences and between the ancestral and observed sequences. Variation in predicted 3-D structural properties of the ancestors mirrored their sequence variation. The observed sequence consensus and ancestral sequences from center-rooted trees were most similar in all predicted attributes. Only for the predicted number of N-glycosylation sites was there a difference between MP and ML methods of reconstruction. Taxon sampling effects were observed only for outgroup-rooted trees, not center-rooted, reflecting the occurrence of several divergent basal sequences. Thus, for sequences exhibiting a radial phylogenetic tree, as does HIV-1, most of the variation in the estimated ancestor arises from the method of rooting the phylogenetic tree. Those investigating the ancestors of genes exhibiting such a radial tree should pay particular attention to alternate rooting methods in order to obtain a representative sample of ancestors.
Duke Biology Box 90338 Durham, NC 27708 Phone: 919-660-7372 Fax: 919-660-7293