Reconstructing genealogies of serial samples under the assumption of a molecular clock using serial-sample UPGMA.


Journal Article

Reconstruction of evolutionary relationships from noncontemporaneous molecular samples provides a new challenge for phylogenetic reconstruction methods. With recent biotechnological advances there has been an increase in molecular sequencing throughput, and the potential to obtain serial samples of sequences from populations, including rapidly evolving pathogens, is fast being realized. A new method called the serial-sample unweighted pair grouping method with arithmetic means (sUPGMA) is presented that reconstructs a genealogy or phylogeny of sequences sampled serially in time using a matrix of pairwise distances. The resulting tree depicts the terminal lineages of each sample ending at a different level consistent with the sample's temporal order. Since sUPGMA is a variant of UPGMA, it will perform best when sequences have evolved at a constant rate (i.e., according to a molecular clock). On simulated data, this new method performs better than standard cluster analysis under a variety of longitudinal sampling strategies. Serial-sample UPGMA is particularly useful for analysis of longitudinal samples of viruses and bacteria, as well as ancient DNA samples, with the minimal requirement that samples of sequences be ordered in time.

Full Text

Duke Authors

Cited Authors

  • Drummond, A; Rodrigo, AG

Published Date

  • December 2000

Published In

Volume / Issue

  • 17 / 12

Start / End Page

  • 1807 - 1815

PubMed ID

  • 11110896

Pubmed Central ID

  • 11110896

Electronic International Standard Serial Number (EISSN)

  • 1537-1719

International Standard Serial Number (ISSN)

  • 0737-4038

Digital Object Identifier (DOI)

  • 10.1093/oxfordjournals.molbev.a026281


  • eng