
2
1 Introduction
Graph width parameters are useful tools for identifying tractable classes of in-
stances for NP-hard problems and designing efficient algorithms for such prob-
lems on these instances. Treewidth and clique-width are two of the most po-
pular graph width parameters. An algorithmic meta-theorem due to Courcelle,
Makowsky, and Rotics [7] states that any problem expressible in the monadic
second-order logic on graphs (MSO1) can be solved in FPT time when parame-
terized by the clique-width of the input graph.3In addition, Courcelle [5] states
that any problem expressible in the monadic second-order logic of graphs with
edge set quantifications (MSO2) can be solved in FPT time when parameterized
by the treewidth of the input graph. Although the class of graphs with bounded
treewidth is a subclass of the class of graphs with bounded clique-width [4], the
MSO2logic on graphs extends the MSO1logic, and there are MSO2properties
like “Ghas a Hamiltonian cycle” that are not MSO1expressible [6]. In addi-
tion, there are problems that are fixed-parameter tractable when parameterized
by treewidth, such as MaxCut,Largest Bond,Longest Cycle,Longest
Path,Edge Dominating Set,Graph Coloring,Clique Cover,Mini-
mum Path Cover, and Minimum Cycle Cover that cannot be FPT when
parameterized by clique-width [12,15,16,17,18], unless FPT = W[1].
For problems that are fixed-parameter tractable concerning treewidth, but
intractable when parameterized by clique-width, the identification of tractable
classes of instances of bounded clique-width and unbounded treewidth becomes
a fundamental quest [11]. In 2016, Dvoˇr´ak, Knop, and Masaˇr´ık [13] showed that
k-Path Cover is FPT when parameterized by the treewidth of the complement
of the input graph. This implies that Hamiltonian Path is FPT when param-
eterized by the treewidth of the complement graph. In 2017, Knop, Kouteck´y,
Masaˇr´ık, and Toufar (WG 2017, [21]) asked about the complexity of deciding
graph problems Πon the complement of Gconsidering a parameter pof G(i.e.,
with respect to p(G)), especially for sparse graph parameters such as treewidth.
In fact, the treewidth of the complement of the input graph, proposed be called
co-treewidth in [11], seems a nice width parameter to deal with dense instances of
problems that are hard concerning clique-width. MaxCut,Clique Cover, and
Graph Coloring are example of problems W[1]-hard concerning clique-width
but FPT-time solvable when parameterized by co-treewidth (see [11]).
The degeneracy of a graph Gis the least ksuch that every induced subgraph
of Gcontains a vertex with degree at most k. Equivalently, the degeneracy of
Gis the least ksuch that its vertices can be arranged into a sequence so that
each vertex is adjacent to at most kvertices preceding it in the sequence. It is
well-known that the degeneracy of a graph is upper bounded by its treewidth;
thus, the class of graphs with bounded treewidth is also a subclass of the class of
graphs with bounded degeneracy. In [11], Duarte, Oliveira, and Souza presented
an algorithmic framework to deal with the degeneracy of the complement graph,
called co-degeneracy, as a parameter.
3Originally this required a clique-width expression as part of the input.