Quality of Gene Order Calculated by Ant Colony Algorithm is Sensitive to Distance Formula

A group of expression levels of gene generated by gene chips is represented by a vector. The smaller the distance of two vectors is, the more similar the two associated genes are. Gene order is the permutation of genes in which similar genes cluster together, which is useful for biologist. And the optimal gene order is equivalent to the shortest route of traveling salesman problem (TSP), in which the associated vectors are used as virtual cities. Ant colony optimization (ACO) is a popular method to solve TSP. In this paper, ACO is applied to calculate gene order. The experiment of this paper shows that the quality of gene order calculated by ACO is sensitive to distance formula of vectors. The contrary fact is shown by this paper that squared Euclidean distance formula generates better quality of gene order than Pearson distance formula, which is used commonly to calculated gene order.