Two new distance based methods for phylogenetic tree reconstruction

Abstract

Phylogenetic tree reconstruction is a fundamental problem in biology. In this paper, we propose two new distance based methods for phylogenetic tree reconstruction. One is called double selecting (DS) method. Gronau and Moran (J. Comput. Biol., 14 (2007): 1–5) proposed a Deepest Least Common Ancestor (DLCA) method for phylogenetic tree reconstruction. Our another method is modified version of DLCA and called mDLCA method. Using some data sets, we study the accuracy of these two methods in obtaining the correct unrooted tree in comparison with that of three other tree-making methods: the neighbor-joining (NJ) method, the unweighted pair group (UPGMA) method and DLCA algorithms. The results show that the DS method and mDLCA method perform same well as the NJ method, and are better than UPGMA and DLCA methods. Especially, the DS method even is slightly better than the NJ method in some cases.