Geometric medians in reconciliation spaces - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Journal Articles Information Processing Letters Year : 2018

Geometric medians in reconciliation spaces

Abstract

In evolutionary biology, it is common to study how various entities evolve together, for example, how parasites coevolve with their host, or genes with their species. Coevolution is commonly modelled by considering certain maps or reconciliations from one evolutionary tree P to another H, all of which induce the same map between the leaf-sets of P and H (corresponding to present-day associations). Recently, there has been much interest in studying spaces of reconciliations, which arise by defining some metric d on the set R(P, H,) of all possible reconciliations between P and H. In this paper, we study the following question: How do we compute a geometric median for a given subset of R(P, H,) relative to d, i.e. an element med 2 R(P, H,) such that X 0 2 d(med , 0)  X 0 2 d(, 0) holds for all 2 R(P, H,)? For a model where so-called host-switches or transfers are not allowed, and for a commonly used metric d called the edit-distance, we show that although the cardinality of R(P, H,) can be super-exponential, it is still possible to compute a geometric median for a set in R(P, H,) in polynomial time. We expect that this result could be useful for computing a summary or consensus for a set of reconciliations (e.g. for a set of suboptimal reconciliations).
Fichier principal
Vignette du fichier
median-ipl-final.pdf (479.53 Ko) Télécharger le fichier
Origin Files produced by the author(s)
Loading...

Dates and versions

hal-01842439 , version 1 (18-07-2018)
hal-01842439 , version 2 (23-12-2018)

Identifiers

Cite

Katharina Huber, Vincent Moulton, Marie-France Sagot, Blerina Sinaimeri. Geometric medians in reconciliation spaces. Information Processing Letters, 2018, 136, pp.96 - 101. ⟨10.1016/j.ipl.2018.04.001⟩. ⟨hal-01842439v2⟩
84 View
118 Download

Altmetric

Share

Gmail Mastodon Facebook X LinkedIn More