Approximating likelihoods under low but variable rates across sites Original Research Article

We present formulae for estimating the likelihood of sequence data in which sites evolve independently at low but variable rates under a tree-indexed Markov process. These methods can be much faster than those presently employed.