Molecular Phylogenetic Trees Are Inferred by Using Minimum Model-based Complexity

摘要

In this paper, we introduce a new concept of data complexity which we call "model-based complexity". Then, this concept is applied to the reconstruction of phylogenetic tree from DNA sequences and a computer simulation is employed to examine the efficiency of this method. The results suggest that minimum model-based complexity method has good asymptotic property compared with traditional maximum likeli-hood method or its modification by Akaike's AIC in the case that the multifurcate tree is considered as a candidate topology of tree and/or longer sequences more than 3000-bp could be used in reconstructing phy-logenetic tree.