A DNA model for solving the hierarchical clustering problem

Abstract

The hierarchical clustering algorithm is one of the clustering algorithms used widely. The typical character of the hierarchical clustering algorithm is to find the “shortest” patterns. But it is not to ensure that to find the global optimal result by adding the shortest pattern each time. In this paper, we use DNA computing to solve this problem. Because during the ligation of the DNA strands, the bio-reaction's character is to combine all the possible single DNA strands into the shortest double DNA strand. We propose a DNA model with the coding strategy and DNA computing algorithm for the hierarchical clustering problem. we use this algorithm to solve the real problem - the retailer's region partition problem. Although we use the computer to simulate the bio-chemical reactions, we will execute this algorithm in the biological lab in the near future.