A practical method for exact computation of subtree prune and regraft distance.

Abstract

Subtree prune and regraft (SPR) is one kind of tree rearrangements that has seen applications in solving several computational biology problems. The minimum number of rooted SPR ((r)SPR) operations needed to transform one rooted binary tree to another is called the (r)SPR distance between the two trees. Computing the (r)SPR distance has been actively studied in recent years. Currently, there is a lack of practical software tools for computing the (r)SPR distance for relatively large trees with large (r)SPR distance. In this article, we present a simple and practical method that computes the exact (r)SPR distance with integer linear programming. By applying this new method on several simulated and real biological datasets, we show that our new method outperforms existing software tools in term of accuracy and efficiency. Our experimental results indicate that our method can compute the exact (r)SPR distance for many large trees with large (r)SPR distance.