Algorithms for the typing of related DNA sequences

Abstract

This article presents an approach to solve the typing problem using algorithms for set covering. Two algorithms to solve this problem were designed and compared. The typing problem consists of distinguishing all sequences using the least possible reagents. For each reagent, their capacity to distinguish between pairs of sequences is represented in a matrix and-then mapped to the set-covering problem. In (J. Lozano Yecora, MS Thesis. Uni. de las Americas, Puebla, Dept. of Comp. Science, 2004), a proof of the equivalence of this problem with the set covering problem is presented; here, the results of an exact algorithm to solve the set-covering problem are presented. This algorithm uses a branch and bound method based on some intuitive properties of the solution, applying recursion as the final resource. Two approximation algorithms were also implemented and tested. One is an improved greedy algorithm and the other is based on dynamic programming. Real as well as randomly created instances were used to test the algorithms.