Weighted minimum-length rearrangement scenarios - LIRMM - Laboratoire d’Informatique, de Robotique et de Microélectronique de Montpellier Access content directly
Conference Papers Year : 2019

Weighted minimum-length rearrangement scenarios


We present the first known model of genome rearrangement with an arbitrary real-valued weight function on the rearrangements. It is based on the dominant model for the mathematical and algorithmic study of genome rearrangement, Double Cut and Join (DCJ). Our objective function is the sum or product of the weights of the DCJs in an evolutionary scenario, and the function can be minimized or maximized. If the likelihood of observing an independent DCJ was estimated based on biological conditions, for example, then this objective function could be the likelihood of observing the independent DCJs together in a scenario. We present an O(n 4)-time dynamic programming algorithm solving the Minimum Cost Parsimonious Scenario (MCPS) problem for co-tailed genomes with n genes (or syntenic blocks). Combining this with our previous work on MCPS yields a polynomial-time algorithm for general genomes. The key theoretical contribution is a novel link between the parsimonious DCJ (or 2-break) scenarios and quadrangulations of a regular polygon. To demonstrate that our algorithm is fast enough to treat biological data, we run it on syntenic blocks constructed for Human paired with Chimpanzee, Gibbon, Mouse, and Chicken. We argue that the Human and Gibbon pair is a particularly interesting model for the study of weighted genome rearrangements.
Fichier principal
Vignette du fichier
LIPIcs-WABI-2019-13.pdf (613.88 Ko) Télécharger le fichier

Dates and versions

lirmm-02410451 , version 1 (13-12-2019)





Pijus Simonaitis, Annie Chateau, Krister M. Swenson. Weighted minimum-length rearrangement scenarios. WABI 2019 - 19th International Workshop on Algorithms in Bioinformatics, Sep 2019, Niagara Falls, NY, United States. pp.13:1-13:17, ⟨10.4230/LIPIcs.WABI.2019.13⟩. ⟨lirmm-02410451⟩
93 View
93 Download



Gmail Facebook X LinkedIn More