Bayes Maximum Likelihood analysis of Harp Seal (Pagophilus groenlandicus) relationships
Harp
Seals were collected from four breeding grounds, the
southern Gulf of St Lawrence (GS),
the Newfoundland and Labrador Ice Front (NL and NL'),
Jan Mayen Island in the Greenland Sea (GS), and the White Sea
(WS).
Several methods of constructing phylogenetic trees from genetic data are available. Bayes Analysis uses conditional probability calculated from an evidence-based modelabout molecular evolution. For example, a 4x4 matrix table of the frequency of all pairwise interchanges of DNA bases can be estimated from the aligned sequences. The likelihood of any tree can be estimated as the probability of the simultaneous occurrence of all necessary molecular changes to produce that tree. The probability of any particular tree is therefore extremely low, but can be estimated. Maximum Likelihood analysis then identifies which of the many unlikely trees is the least unlikely, that is, maximally likely. ML identifies six major lineages (clades) (A - F) and within these up to five sub-lineages (A1 - A5). The statistical confidence in these groups is measured by posterior probabilities.
Several methods of constructing phylogenetic trees from genetic data are available. Bayes Analysis uses conditional probability calculated from an evidence-based modelabout molecular evolution. For example, a 4x4 matrix table of the frequency of all pairwise interchanges of DNA bases can be estimated from the aligned sequences. The likelihood of any tree can be estimated as the probability of the simultaneous occurrence of all necessary molecular changes to produce that tree. The probability of any particular tree is therefore extremely low, but can be estimated. Maximum Likelihood analysis then identifies which of the many unlikely trees is the least unlikely, that is, maximally likely. ML identifies six major lineages (clades) (A - F) and within these up to five sub-lineages (A1 - A5). The statistical confidence in these groups is measured by posterior probabilities.