Improved parameterized algorithms for network query problems

Ron Y. Pinter, Hadas Shachnai, Meirav Zehavi

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

3 Scopus citations

Abstract

In the PARTIAL INFORMATION NETWORK QUERY (PINQ) problem, we are given a host graph H, and a pattern P whose topology is partially known. We seek a subgraph of H that resembles P. PINQ is a generalization of Subgraph Isomorphism, where the topology of P is known, and Graph Motif, where the topology of P is unknown. This generalization has important applications to bioinformatics, since it addresses the major challenge of analyzing biological networks in the absence of certain topological data. In this paper, we use a non-standard part-algebraic/part-combinatorial hybridization strategy to develop an exact parameterized algorithm as well as an FPT-approximation scheme for PINQ, allowing near resemblance between H and P. We thus unify and significantly improve previous results related to network queries.

Original languageEnglish
Title of host publicationParameterized and Exact Computation - 9th International Symposium, IPEC 2014, Revised Selected Papers
EditorsMarek Cygan, Pinar Heggernes
PublisherSpringer Verlag
Pages294-306
Number of pages13
ISBN (Electronic)9783319135236
DOIs
StatePublished - 1 Jan 2014
Externally publishedYes
Event9th International Symposium on Parameterized and Exact Computation, IPEC 2014 - Wroclaw, Poland
Duration: 10 Sep 201412 Sep 2014

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume8894
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference9th International Symposium on Parameterized and Exact Computation, IPEC 2014
Country/TerritoryPoland
CityWroclaw
Period10/09/1412/09/14

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science

Fingerprint

Dive into the research topics of 'Improved parameterized algorithms for network query problems'. Together they form a unique fingerprint.

Cite this