Navigation Banner
 

     Full model name: Tamura and Nei, 1993

     Model type: Nucleotide

     Model options:

  • Local. Rates are homogeneous (same for all sites in the data). Each branch has 3 parameters (transversion rate and 2 transition rates, one for purines, the other for pyrimidines).
  • Global. Rates are homogeneous (same for all sites in the data). Each branch has 1 parameter (transversion rate). 2 transition/transversion ratios are shared by all branches in the tree.
  • Global+Variation Each branch has 1 parameter (transversion rate). 2 transition/transversion ratios are shared by all branches in the tree. Different sites may have different rates. The user specifies how many different rate classes there should be 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:

  • Local. a - transversion rate, b - purine transition rate, c - pyrimidine transition rate; (for each branch)
  • Global. a - transversion rate (for each branch); R1 - transition/transversion ratio for purines; R2 - transition/transversion ratio for pyrimidines; (shared by all branches in the tree).
  • Rate Variation (+HM). Same as in the global setting plus the appropriate distribution parameters.

 

    Rate Matrix:

  • Local.

  • Global, Rate Variation (+HM):

    with the vector representing nucleotide frequencies (proportions) from the data. '*' is defined to be the negative of the sum of all other row elements.

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