Median clouds and a fast transposition median solver

Paper in proceeding, 2009

The median problem seeks a permutation whose total distance to a given set of permutations (the base set) is minimal. This is an important problem in comparative genomics and has been studied for several distance measures such as reversals. The transposition distance is less relevant biologically, but it has been shown that it behaves similarly to the most important biological distances, and can thus give important information on their properties. We have derived an algorithm which solves the transposition median problem, giving all transposition medians (the median cloud). We show that our algorithm can be modified to accept median clouds as elements in the base set and briefly discuss the new concept of median iterates (medians of medians) and limit medians, that is the limit of this iterate.