Finding a chain graph in a bipartite permutation graph☆
Introduction
Given a base graph G and a pattern graph H, Subgraph Isomorphism (SGI for short) asks whether G contains a subgraph isomorphic to H, where a subgraph is a graph obtained by removing some edges and vertices. The problem SGI is in NP and generalizes many NP-complete problems such as Hamiltonicity, Clique, and Bandwidth. Thus SGI is NP-complete in general [2]. The complexity of SGI is studied in many aspects including the parameterized complexity and graph classes. In this paper, we study SGI by restricting input graphs to be in some graph classes. For studies in the parameterized complexity of SGI, see the recent papers by Marx and Pilipczuk [7], Jansen and Marx [4], and the references therein.
Since the problem SGI immediately becomes NP-complete if we allow the input graph class to contain all unions of disjoint paths or all unions of disjoint cliques [1], an easy NP-hardness reduction works for most of graph classes such as forests and cographs. Kijima et al. [5] thus studied a restricted version of SGI that they call Spanning Subgraph Isomorphism (SSGI, for short), where the base and pattern graphs are connected and have the same number of vertices. They showed that SSGI is NP-complete even for bipartite permutation graphs, proper interval graphs, and trivially perfect graphs. On the other hand, they also showed that SGI, the problem without the restrictions, is polynomial-time solvable for chain graphs, cochain graphs, and threshold graphs.
Recently, Konagaya et al. [6] have narrowed the complexity gap by showing that SGI is polynomial-time solvable if the base graphs are proper interval graphs (or the even larger class of chordal graphs) and the pattern graphs are cochain graphs, or if the base graphs are trivially perfect graphs and the pattern graphs are threshold graphs. The complexity of the case where the base graphs are bipartite permutation graphs and the pattern graphs are chain graphs remained unsettled.
In this paper, we study the unsettled case and show that it is polynomial-time solvable. That is, we show that SGI is polynomial-time solvable if the base graphs are bipartite permutation graphs and the pattern graphs are chain graphs.
Section snippets
Preliminaries
Let and be graphs. We say that H is subgraph-isomorphic to G if there exists an injective adjacency-preserving map η from to ; that is, if and holds for each . We call such a map η a subgraph-isomorphism from H to G. We call G and H the base graph and the pattern graph, respectively. Now SGI can be formally stated as follows: Problem SGI Instance: A pair of graphs G and H. Question: Is H subgraph-isomorphic to G?
Let be a graph. For
Finding a connected chain graph in a connected 2-layer chain graph
Let be a connected 2-layer chain graph with , , and . Let be a connected chain graph with and .
Note that since both G and H are connected bipartite graphs, a subgraph-isomorphism η from H to G satisfies either or . Also, implies . In the rest of this section, we assume that we correctly guessed that and since we can perform the algorithm twice.
Concluding remarks
As a final remark, we now show an upper bound of the running time, where n is the number of vertices in the base graph. Konagaya et al. [6] showed that by solving instances of SGI with connected 2-layer chain graphs as the base graphs and connected chain graphs as the pattern graphs, one can solve the problem with bipartite permutation graphs as the base graphs and chain graphs as the pattern graphs. They can produce each instance in linear time. In our algorithm, given a produced
References (8)
- et al.
Subgraph isomorphism in graph classes
Discrete Math.
(2012) - et al.
Polynomial-time algorithms for Subgraph Isomorphism in small graph classes of perfect graphs
Discrete Appl. Math.
(2016) Induced subgraph isomorphism for cographs is NP-complete
- et al.
Computers and Intractability: A Guide to the Theory of NP-Completeness
(1979)
Cited by (5)
Labelled well-quasi-order for permutation classes
2022, Combinatorial TheorySubgraph Isomorphism on Graph Classes that Exclude a Substructure
2020, AlgorithmicaSubgraph Isomorphism on Graph Classes that Exclude a Substructure
2019, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)