Navigation Banner
 

     Full model name: Dayhoff Empricial model.

     Model type: Amino acids.

     Model options:

  • Fixed rates. . Rates are homogeneous (same for all sites in the data).
  • Rate Variation. Different sites may have different rates. The user specifies how many different rate classes there should and selects one of the available distributions to be sampled. Distribution parameters are estimated.
  • Rate Variation+HM. Similar to Rate Variation except that rates at adjacent sites are correlated via a Hidden Markov Model

     Independent Parameters: t (for each branch of the tree) and, if rate variation is selected as an option, the corresponding distribution/hmm parameters.

     Rate Matrix (1/1000 * the following matrix):

   '*' is defined to be the negative of the sum of all other row elements.

    Each column is multiplied by the equilibrium frequency of the amino acid indicated above that column, and those frequencies are drawn from the data below:

 
Sergei L. Kosakovsky Pond and Spencer V. Muse, 1997-2002