On a discretizable subclass of instances of the molecular distance geometry problem

  • Carlile Lavor
  • , Leo Liberti
  • , Antonio Mucherino
  • , Nelson Maculan

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

The molecular distance geometry problem can be formulated as the problem of finding an immersion in R3 of a given undirected, nonnegatively weighted graph G. In this paper, we discuss a set of graphs G for which the problem may also be formulated as a combinatorial search in discrete space. This is theoretically interesting as an example of "combinatorialization" of a continuous nonlinear problem. It is also algorithmically interesting because the natural combinatorial solution algorithm performs much better than a global optimization approach on the continuous formulation. We present a Branch and Prune algorithm which can be used for obtaining a set of positions of the atoms of protein conformations when only some of the distances between the atoms are known.

Original languageEnglish
Title of host publication24th Annual ACM Symposium on Applied Computing, SAC 2009
PublisherAssociation for Computing Machinery (ACM)
Pages804-805
Number of pages2
ISBN (Print)9781605581668
DOIs
Publication statusPublished - 1 Jan 2009
Event24th Annual ACM Symposium on Applied Computing, SAC 2009 - Honolulu, HI, United States
Duration: 8 Mar 200912 Mar 2009

Publication series

NameProceedings of the ACM Symposium on Applied Computing

Conference

Conference24th Annual ACM Symposium on Applied Computing, SAC 2009
Country/TerritoryUnited States
CityHonolulu, HI
Period8/03/0912/03/09

Keywords

  • Combinatorialization
  • Distance geometry
  • Protein backbone
  • Protein molecules
  • Undirected weighted graph

Fingerprint

Dive into the research topics of 'On a discretizable subclass of instances of the molecular distance geometry problem'. Together they form a unique fingerprint.

Cite this