Finding large $H$-colorable subgraphs in hereditary graph classes

We study the \textsc{Max Partial $H$-Coloring} problem: given a graph $G$, find the largest induced subgraph of $G$ that admits a homomorphism into $H$, where $H$ is a fixed pattern graph without loops. Note that when $H$ is a complete graph on $k$ vertices, the problem reduces to finding the largest induced $k$-colorable subgraph, which for $k=2$ is equivalent (by complementation) to \textsc{Odd Cycle Transversal}. We prove that for every fixed pattern graph $H$ without loops, \textsc{Max Partial $H$-Coloring} can be solved: $\bullet$ in $\{P_5,F\}$-free graphs in polynomial time, whenever $F$ is a threshold graph; $\bullet$ in $\{P_5,\textrm{bull}\}$-free graphs in polynomial time; $\bullet$ in $P_5$-free graphs in time $n^{\mathcal{O}(ω(G))}$; $\bullet$ in $\{P_6,\textrm{1-subdivided claw}\}$-free graphs in time $n^{\mathcal{O}(ω(G)^3)}$. Here, $n$ is the number of vertices of the input graph $G$ and $ω(G)$ is the maximum size of a clique in~$G$. Furthermore, combining the mentioned algorithms for $P_5$-free and for $\{P_6,\textrm{1-subdivided claw}\}$-free graphs with a simple branching procedure, we obtain subexponential-time algorithms for \textsc{Max Partial $H$-Coloring} in these classes of graphs. Finally, we show that even a restricted variant of \textsc{Max Partial $H$-Coloring} is $\mathsf{NP}$-hard in the considered subclasses of $P_5$-free graphs, if we allow loops on $H$.