This paper was converted on www.awesomepapers.org from LaTeX by an anonymous user.
Want to know more? Visit the Converter page.

A Note Related to Graph Theory

Jinfeng Li
Abstract

This article foucuses on (P3P2,K4)(P_{3}\cup P_{2},K_{4})-free graph. In this paper, we prove that if G is (P3P2,K4)(P_{3}\cup P_{2},K_{4})-free, then χ(G)7\chi(G)\leq 7. We then use this result to obtain the upper bound of order and chromatic number of (4K1,P3P2¯,Kω)(4K_{1},\overline{P_{3}\cup P_{2}},K_{\omega})-free graph .

Keywords: coloring; chromatic number; clique number; χ\chi-binding function; P3P2P_{3}\cup P_{2}-free; K4K_{4}-free.

1 Introduction

A graph GG consists of its vertex set V(G)V(G) and edge set E(G)V(G)×V(G)E(G)\subseteq V(G)\times V(G). The order of GG, denoted by nn, is the size of V(G)V(G). The complement of a graph GG, denoted by G¯\overline{G} or coGco-G, is a graph with the same vertex set while whose edge set consists of the edges not present in E(G)E(G).

All graphs in this paper are finite and have no loops or parallel edges. A graph GG contains HH if HH is isomorphic to an induced subgraph of GG. If H1H_{1} is isomorphic to H2H_{2}, then we say H1H2H_{1}\simeq H_{2}. For a collection of graphs Ht,t=1,2,,nH_{t},t=1,2,...,n, GG is (H1,H2,,Hn)(H_{1},H_{2},...,H_{n})-free if it does not contain Ht,t=1,2,,nH_{t},t=1,2,...,n. In this paper, if GG does not contain P2P_{2}, we will say that GG is edge-free.

Path and cycle on nn vertex is denoted by PnP_{n} and CnC_{n}, respectively. The complete graph on that has nn vertex is denoted bu KnK_{n}. We use GHG\cup H to denote the disjoint union of GG and HH.

A kk-coloring of a graph GG is a function ϕ:V(G){1,,k}\phi:V(G)\rightarrow\{1,...,k\} such that ϕ(u)ϕ(v)\phi(u)\neq\phi(v) whenever uu and vv are adjacent in GG. We say that GG is kk-colorable if GG admits a kk-coloring. The chromatic number of GG is denoted by χ(G)\chi(G) which represents the minimum positive integer kk such that GG is kk-colorable. The clique number is denoted by ω(G)\omega(G) which represents the size of the largest clique in GG.

A family 𝔾\mathbb{G} of graphs is said to be χ\chi-bounded if there exists a function f such that for every graph G𝔾G\in\mathbb{G} and every induced subgraph HH of GG it holds that χ(H)f(ω(H))\chi(H)\leq f(\omega(H)). The function ff is called a χ\chi-binding function for GG. The class of perfect graphs (a graph GG is perfect if for every induced subgraph HH of GG it holds that χ(H)=ω(H))\chi(H)=\omega(H)), for instance, is a χ\chi-bounded family with χ\chi-binding function f(x)=xf(x)=x. Therefore, χ\chi-boundedness is a generalization of perfection. The notion of χ\chi-bounded families was introduced by Gyárfás who make the following conjecture.

conjecture 1.1.

[3] For every forest T, the class of TT-free graphs is χ\chi-bounded.

For ω=3\omega=3, Esperet, Lemoine, Maffray and Morel[2] obtained the optimal bound on the chromatic number: every (P5,K4)(P_{5},K_{4})-free graph is 5-colorable. Serge Gaspers and Shenwei Huang[6] obtained the optimal bound of the chromatic number: every (2P2,K4)(2P_{2},K_{4})-free graph is 4-colorable. Karthick and S. Mishra obtained the optimal bound of the chromatic number: every (P6,diamond,K4)(P_{6},diamond,K_{4})-free graph is 6-colorable. Rui Li,Jinfeng Li and Di Wu claimed that [5] (P3P2,K4)(P_{3}\cup P_{2},K_{4})-free graph is 99-colorable. We follow the methods in this paper and sharper the bound to 77.

Partition V(G)V(G) into two following parts:

  • D1:={xV(G)|ω(GN(x))2}D_{1}:=\{x\in V(G)|\omega(G-N(x))\leq 2\}

  • D2:=GD1D_{2}:=G-D_{1}

Let C=u1v1v2v3u3u2C=u_{1}v_{1}v_{2}v_{3}u_{3}u_{2} be a 66-hole. We use codoninoco-donino to denote a graph obtained from CC by connecting edges v1v3v1v3 and u1u3u1u3.

Let C=u1v1v2v3u3u2C=u_{1}v_{1}v_{2}v_{3}u_{3}u_{2} be a 66-hole. We use coAco-A to denote a graph obtained from CC by connecting edges v1v3v_{1}v_{3} ,u1u3u_{1}u_{3} and u1v3u_{1}v_{3}.

Our proof includes three major theorem:

theorem 1.1.

If G[D1]G[D_{1}] contains P2P1P_{2}\cup P_{1}, then χ(G)7\chi(G)\leq 7.

theorem 1.2.

If GG contains codominoco-domino or coAco-A, then χ(G)7\chi(G)\leq 7.

theorem 1.3.

If GG is (codomino,coA)(co-domino,co-A)-free and G[D2]G[D_{2}] is not an induced subgraph of K3K_{3}, then χ(G)7\chi(G)\leq 7.

Finally, we obtain a theorem for (4K1,P3P2¯)(4K_{1},\overline{P_{3}\cup P_{2}})-free graphs:

theorem 1.4.

If GG is (4K1,P3P2¯)(4K_{1},\overline{P_{3}\cup P_{2}})-free with clique number ω\omega and order n, then n7ωn\leq 7\omega and χ(G)4ω\chi(G)\leq 4\omega.

2 Structure Around K3K_{3}

Suppose there is a K3K_{3} in GG and its vertex set is {v1,v2,v3}\{v_{1},v_{2},v_{3}\}. Naturally, we can divide V(G)V(G) into following three sets:

  • A0:={v|v is not adjacent to any vertex in {v1,v2,v3}}A_{0}:=\{v|v\text{ is not adjacent to any vertex in }\{v_{1},v_{2},v_{3}\}\}.

  • A1:={v|v is adjacent to one vertex in {v1,v2,v3} only}A_{1}:=\{v|v\text{ is adjacent to one vertex in }\{v_{1},v_{2},v_{3}\}\text{ only}\}.

  • A2:={v|v is adjacent to two vertex in {v1,v2,v3} only}A_{2}:=\{v|v\text{ is adjacent to two vertex in }\{v_{1},v_{2},v_{3}\}\text{ only}\}.

we need to combine A0A_{0} and A1A_{1} and divide it into three disjoint sets: B1,B2,B3B_{1},B_{2},B_{3}. The set B1B_{1} includes vertices that is only adjacent to u1u_{1} and vertices in A0A_{0}. The set Bj,j=2,3B_{j},j=2,3 includes vertices that are only adjacent to uju_{j}. Since G[Bi]G[B_{i}] is anticomplete to an edge in G[{v1,v2,v3}]G[\{v_{1},v_{2},v_{3}\}], G[Bi],i=1,2,3G[B_{i}],i=1,2,3 is P3P_{3}-free.

A2{v1,v2,v3}A_{2}\cup\{v_{1},v_{2},v_{3}\} can be partitioned in to (N(vi)N(i1)){vi+1}(N(v_{i})\cap N(_{i-1}))\cup\{v_{i+1}\}(mod3) for i=1,2,3i=1,2,3. It is easy to obtain that χ((N(vi)N(i1)){vi+1})1\chi((N(v_{i})\cap N(_{i-1}))\cup\{v_{i+1}\})\leq 1(mod3) for i=1,2,3i=1,2,3.

3 main theorem

We first introduce a lemma provided by Rui Li, Jinfeng Li and Di Wu.

lemma 3.1.

[5] If GG contains P2K3P_{2}\cup K_{3}, then χ(G)6\chi(G)\leq 6.

Define D1:={x|ω(GN(x))2}D_{1}:=\{x|\omega(G-N(x))\leq 2\} and D2:=GD1D_{2}:=G-D_{1}.

claim 3.2.

For any two nonadjacent vertex y1,y2y_{1},y_{2} in D1D_{1}, ω(G[N(y1)N(y2)])1\omega(G[N(y_{1})-N(y_{2})])\leq 1 and ω(G[N(y2)N(y1)])1\omega(G[N(y_{2})-N(y_{1})])\leq 1.

Proof. If there is an edge in G[N(y1)N(y2)]G[N(y_{1})-N(y_{2})](or G[N(y2)N(y1)]G[N(y_{2})-N(y_{1})]), then ω(G[N(y1)N(y2)])=3\omega(G[N(y_{1})-N(y_{2})])=3(or ω(G[N(y2)N(y1)])=3\omega(G[N(y_{2})-N(y_{1})])=3), which contradicts definition of y1y_{1}(or y2y_{2}).∎

We fisrt introduce a lemma to help us slightly determine the structure of G[D1]G[D_{1}].

theorem.

(1.1) If G[D1]G[D_{1}] contains P2P1P_{2}\cup P_{1}, then χ(G)7\chi(G)\leq 7.

Proof. Suppose G[{v1,v2}{v3}]P2P1G[\{v_{1},v_{2}\}\cup\{v_{3}\}]\simeq P_{2}\cup P_{1}, M(v1,v2)=G{v1,v2}(N(v1)N(v2))M(v_{1},v_{2})=G-\{v_{1},v_{2}\}-(N(v_{1})\cup N(v_{2})). we can divide GG into {v1,v2}(N(v1)N(v2))M(v1,v2)\{v_{1},v_{2}\}\cup(N(v_{1})\cup N(v_{2}))\cup M(v_{1},v_{2}). Since GG is (P3P2,P2K3)(P_{3}\cup P_{2},P_{2}\cup K_{3})-free, G[M(v1,v2)]G[M(v_{1},v_{2})] is unions of clique and ω(M(v1,v2))2\omega(M(v_{1},v_{2}))\leq 2. Therefore, {v1,v2}M(v1,v2)2\{v_{1},v_{2}\}\cup M(v_{1},v_{2})\leq 2.

We will prove χ(N(v1)N(v2))5\chi(N(v_{1})\cup N(v_{2}))\leq 5 and hence χ(G)7\chi(G)\leq 7. N(v1)N(v2)N(v_{1})\cup N(v_{2}) can be divided into N(v1)N(v3)N(v_{1})-N(v_{3}), (N(v1)N(v3))N(v2)(N(v_{1})\cap N(v_{3}))-N(v_{2}), N(v2)N(v3)N(v_{2})-N(v_{3}), (N(v2)N(v3))N(v1)(N(v_{2})\cap N(v_{3}))-N(v_{1}) and N(v1)N(v2)N(v_{1})\cap N(v_{2}). We apply claim 3.3 to prove that the clique number of first four sets are at most 1. If the last set has edge, then ω(G)4\omega(G)\geq 4. Therefore, χ(N(v1)N(v2))5\chi(N(v_{1})\cup N(v_{2}))\leq 5. ∎

observation 1.

If G[D1]G[D_{1}] is P2P1P_{2}\cup P_{1}-free, then G[D1]¯\overline{G[D_{1}]} is P3P_{3}-free.

Now we introduce a lemma which is important in the proof of theorem1.2.

claim 3.3.

Suppose V(G)V(G) can be partitioned into three nonempty subsets V1V_{1} , V2V_{2} and V3V_{3} such that G[V1]G[V_{1}] and G[V2]G[V_{2}] are (K3,P3)(K_{3},P_{3})-free graph, G[V3]G[V_{3}] is a stable set , G[V1V2]G[V_{1}\cup V_{2}] is K3K_{3}-free , G[VjV3]G[V_{j}\cup V_{3}] is (K3,P3)(K_{3},P_{3})-free graph for j=1,2j=1,2 and any vertex vv in Vi,i=1,2V_{i},i=1,2, M(v)ViM(v)-V_{i} has no edge. Furthermore, for any vertex vV3v\in V_{3}, if N(v)ViN(v)\cap V_{i}\neq\emptyset, then N(v)Vi={xVi|x is not complete to N(v)N3i},i=1,2N(v)\cap V_{i}=\{x\in V_{i}|x\text{ is not complete to }N(v)\cap N_{3-i}\},i=1,2. If G[V1]G[V_{1}] has at most one edge, then χ(G)3\chi(G)\leq 3.

Proof.

If V1V_{1} has no edge, then randomly select one vertex vV1v\in V_{1}. We partition V(G)V(G) accoding to {v}\{v\}, that is V(G)={v}N(v)M(v)V(G)=\{v\}\cup N(v)\cup M(v). |N(v)V3|1|N(v)\cap V_{3}|\leq 1. Now we partition V(G)V(G) into three stable set: (N(v)V3)(V1M(v))(N(v)\cap V_{3})\cup(V_{1}\cap M(v)), N(v)V2N(v)\cap V_{2} and v(M(v)V1){v}\cup(M(v)-V_{1}). (N(v)V3)(V1M(v))(N(v)\cap V_{3})\cup(V_{1}\cap M(v)) is stable set as G[V3V1]G[V_{3}\cup V_{1}] is P3P_{3}-free. According to definition, the rest two sets are stable sets.

If V1V_{1} has one edge, then set one vertex of the vertex set of that edge vv. We partition V(G)V(G) into three part: V(G)={v}N(v)M(v)V(G)=\{v\}\cup N(v)\cup M(v). It is obvious that N(v)V2N(v)\subseteq V_{2}, otherwise G[V1V2]G[V_{1}\cup V_{2}] induces an P3P2P_{3}\cup P_{2} or K3P2K_{3}\cup P_{2}. Therefore, as G[V1V2]G[V_{1}\cup V_{2}] is K3K_{3}-free, we can conclude that χ(N(v))1\chi(N(v))\leq 1. χ(M(v))χ(M(v)V1)+χ(M(v)V1)2\chi(M(v))\leq\chi(M(v)\cap V_{1})+\chi(M(v)-V_{1})\leq 2 and hence χ(G)χ(N(v))+χ({v}M(v))3\chi(G)\leq\chi(N(v))+\chi(\{v\}\cup M(v))\leq 3. ∎

observation 2.

If GG contains codominoco-domino, then according to section 2.1, V(G)V(G) can be partitioned into B1,B2,B3B_{1},B_{2},B_{3} and A2A_{2}.

lemma 3.4.

If GG contains codominoco-domino, then χ(G)7\chi(G)\leq 7.

Proof.

[Uncaptioned image]

ω(G[B2])1\omega(G[B_{2}])\leq 1. Otherwise there exists an edge in G[B2]G[B_{2}]. We call the edge G[{y1,y2}]G[\{y_{1},y_{2}\}]. If y1u2y_{1}\sim u_{2}, then G[{v1,v3}{y2,y1,u2}]P3P2G[\{v_{1},v_{3}\}\cup\{y_{2},y_{1},u_{2}\}]\simeq P_{3}\cup P_{2} or G[{v1,v3}{y2,y1,u2}]K3P2G[\{v_{1},v_{3}\}\cup\{y_{2},y_{1},u_{2}\}]\simeq K_{3}\cup P_{2}(depending on whether y2y_{2} is adjacent to u2u_{2}). As a consequence, both y1y_{1} and y2y_{2} are not adjacent to u2u_{2}. However, y1y_{1} is complete to {u1,u3}\{u_{1},u_{3}\}, which can be proven through making use of P2P3P_{2}\cup P_{3}-free. Similarly, y2y_{2} is complete to {u1,u3}\{u_{1},u_{3}\}. As a consequence, G[{y1,y2,u3,u1}]K4G[\{y_{1},y_{2},u_{3},u_{1}\}]\simeq K_{4}, which contradicts the fact that ω(G)3\omega(G)\leq 3. As a consequence, χ(B2)1\chi(B_{2})\leq 1.

For any xB2x\in B_{2}, N(x)N(x) must be one of four following sets: {u1,u2}\{u_{1},u_{2}\},{u2,u3}\{u_{2},u_{3}\},{u2}\{u_{2}\} and {u3,u1}\{u_{3},u_{1}\}. We partition B2B_{2}({u1,u2}\{u_{1},u_{2}\}) into C1C_{1}({u2,u3}\{u_{2},u_{3}\}),C2C_{2}({u2,u3}\{u_{2},u_{3}\}),C3C_{3}({u2}\{u_{2}\}) and C4C_{4}({u3,u1}\{u_{3},u_{1}\}) accoding to their neigbors in {u1,u2,u3}\{u_{1},u_{2},u_{3}\}. Trivially , CiCj=C_{i}\cap C_{j}= for ij{1,2,3,4}i\neq j\in\{1,2,3,4\}.

If C1C_{1}\neq\emptyset, then there is no edge in G[B3]G[B_{3}]. Suppose {y1,y2}E(B3)\{y_{1},y_{2}\}\in E(B_{3}) and xC1x\in C_{1}, we are going to prove yi,i=1,2y_{i},i=1,2 are complete to {v1,x}\{v_{1},x\} and anticomplete to {u2,u3}\{u_{2},u_{3}\} and hence G[{y1,y2,u1,x}]K4G[\{y_{1},y_{2},u_{1},x\}]\simeq K_{4}, which contradicts ω(G)<4\omega(G)\textless 4. If y1y_{1} is adjacent to u2u_{2}, then G[{y1,u3,u2}{v2,v1}]G[\{y_{1},u_{3},u_{2}\}\cup\{v_{2},v_{1}\}] is isomorphic to P2P3P_{2}\cup P_{3} or P2K3P_{2}\cup K_{3}. Symmetrically, y2y_{2} is anticomplete to {u2,u3}\{u_{2},u_{3}\}. As consequence, y1y_{1} must be complete to {u1,x}\{u_{1},x\}, or G[{u2,u1,x,v2,v3,y1}]G[\{u_{2},u_{1},x,v_{2},v_{3},y_{1}\}] will induce P3P2P_{3}\cup P_{2}. Symmetrically, y2y_{2} is complete to {x,u1}\{x,u_{1}\}. As consequence, we get the contradiction that G[{y1,y2,u1,x}]K4G[\{y_{1},y_{2},u_{1},x\}]\simeq K_{4}. Therefore, ω(B3)1\omega(B_{3})\leq 1.

If C2C_{2}\neq\emptyset, then there is no edge in G[B1A0]G[B_{1}-A_{0}], which is similar to the proof for C1.C_{1}\neq\emptyset.

From dicussion above, if C1C2C_{1}\cup C_{2}\neq\emptyset, then χ(B1B2B3)4\chi(B_{1}\cup B_{2}\cup B_{3})\leq 4 and hence χ(G)χ(B1B2B3)+χ(A2{v1,v2,v3})4+3=7\chi(G)\leq\chi(B_{1}\cup B_{2}\cup B_{3})+\chi(A_{2}\cup\{v_{1},v_{2},v_{3}\})\leq 4+3=7.

Suppose C1C2=C_{1}\cup C_{2}=\emptyset and xC3x\in C_{3}. Furthermore, |C3|1|C_{3}|\leq 1. Otherwise, G[{u2}{v1,v3}C3]G[\{u_{2}\}\cup\{v_{1},v_{3}\}\cup C_{3}] induces P3P2P_{3}\cup P_{2} or P2K3P_{2}\cup K_{3}. According to the definition, A2A_{2} can be partitioned into N(v1)N(v2)N(v_{1})\cap N(v_{2}), N(v1)N(v3)N(v_{1})\cap N(v_{3}) and N(v2)N(v3)N(v_{2})\cap N(v_{3}).

Suppose there is a vertex in (N(v2)N(v3)(N(v_{2})\cap N(v_{3}) has only one neighbor in {u1,u2,u3}\{u_{1},u_{2},u_{3}\}. It is not difficult to obtain that the only neighbor is v2v_{2}. If we exchange {v1,v2,v3}\{v_{1},v_{2},v_{3}\} and {u1,u2,u3}\{u_{1},u_{2},u_{3}\} in codominoco-domino, then we can prove χ(G)7\chi(G)\leq 7 in the same way as what we did when C1C2C_{1}\cup C_{2}\neq\emptyset.

Suppose evrey vertex in A2N(v2)A_{2}\cap N(v_{2}) has two neighbors in {u1,u2,u3}\{u_{1},u_{2},u_{3}\}.

Let C=u1v1v2v3u3u2C=u_{1}v_{1}v_{2}v_{3}u_{3}u_{2} be a codominoco-domino and {u,x}\{u,x\} be two vertex outside CC. We use co-donino1 to denote a family of graphs obtained from {u,x}C\{u,x\}\cup C by connecting edges uu1,uu2,uv1,uv2,xv2,xu2uu_{1},uu_{2},uv_{1},uv_{2},xv_{2},xu_{2} and xuxu can be connected or disconnected.

Let C=u1v1v2v3u3u2C=u_{1}v_{1}v_{2}v_{3}u_{3}u_{2} be a codominoco-domino and {u,x}\{u,x\} be two vertex outside CC. We use co-domino2 to denote a graph obtained from CC by connecting edges uu1,uu2,uv1,uv3,xv2,xu2uu_{1},uu_{2},uv_{1},uv_{3},xv_{2},xu_{2} and xuxu. Suppose uN(v2)N(v1)u\in N(v_{2})\cap N(v_{1}). The following case shows that if uu is complete to {u1,u2}\{u_{1},u_{2}\} or complete to {u2,u3,x}\{u_{2},u_{3},x\}, then χ(G)7\chi(G)\leq 7. In other words, if GG contains co-domino1 or co-domino2, then χ(G)7\chi(G)\leq 7.

[Uncaptioned image]
[Uncaptioned image]

Recalling that G[B3]G[B_{3}] is anticomplete to u2u_{2} , B3{u3}B_{3}-\{u_{3}\} is anticomplete to u2u_{2}.

case1: GG contains codomino1co-domino1 or codomino2co-domino2.

Suppose GG contains codomino1co-domino1.

G[B3]G[B_{3}] contains at most one edge. Otherwise, suppose G[B3]G[B_{3}] contains more than one edge. Since G[B3]G[B_{3}] is P3P_{3} -free and complete to {u1}\{u_{1}\} and anticomplete to {u3}\{u_{3}\}, there exists an mm such that these edges are isomorphic to mK2,m>1mK_{2},m\textgreater 1 and uu has at most one neighbor in every edges or G[{u,u1}B3]G[\{u,u_{1}\}\cup B_{3}] induces K4K_{4}. However, m>1m\textgreater 1 leads to a contradiction that G[{B3N(u)}{u,u2}]G[\{B_{3}-N(u)\}\cup\{u,u_{2}\}] induces P3P2P_{3}\cup P_{2}.

Let G[B1A0]=V1,G[B3]=V2,G[A0]=V3G[B_{1}-A_{0}]=V_{1}^{\prime},G[B_{3}]=V_{2}^{\prime},G[A_{0}]=V_{3}^{\prime}. According to claim3.3, χ(B3B1)3\chi(B_{3}\cup B_{1})\leq 3. Therefore, χ(G)χ(B1B3)+χ(B2)+χ(A2)3+1+3=7\chi(G)\leq\chi(B_{1}\cup B_{3})+\chi(B_{2})+\chi(A_{2})\leq 3+1+3=7.

Suppose GG is codomino1co-domino1-free and contains codomino2co-domino2.

G[B1A0]G[B_{1}-A_{0}] contains at most one edge. Since G[B1A0]G[B_{1}-A_{0}] is P3P_{3} -free and complete to {u1}\{u_{1}\} and anticomplete to {u3}\{u_{3}\}, this union of edges is isomorphic to mK2,m>1mK_{2},m\textgreater 1 for some mm and uu has at most one neighbor in every edge or G[{u,x}B3]G[\{u,x\}\cup B_{3}] induces K4K_{4}. However, m>1m\textgreater 1 leads to a contradiction that G[{B3N(u)}{u,u2}]G[\{B_{3}-N(u)\}\cup\{u,u_{2}\}] induces P3P2P_{3}\cup P_{2}.

Let G[B1A0]=V1,G[B3]=V2,G[A0]=V3G[B_{1}-A_{0}]=V_{1}^{\prime},G[B_{3}]=V_{2}^{\prime},G[A_{0}]=V_{3}^{\prime}, then according to claim3.3, χ(B3B1)3\chi(B_{3}\cup B_{1})\leq 3. Furthermore, χ(G)χ(B1B3)+χ(B2)+χ(A2)3+1+3=7\chi(G)\leq\chi(B_{1}\cup B_{3})+\chi(B_{2})+\chi(A_{2})\leq 3+1+3=7. Let Let C=u1v1v2v3u3u2C=u_{1}v_{1}v_{2}v_{3}u_{3}u_{2} be a codominoco-domino and {u,x}\{u,x\} be two vertex outside CC. We use co-domino3 to denote a graph obtained from CC by connecting edges uu1,uu2,uv1,uv3,xv2uu_{1},uu_{2},uv_{1},uv_{3},xv_{2} and xu2xu_{2}.

[Uncaptioned image]

Suppose uN(v1)N(v2)u\in N(v_{1})\cap N(v_{2}) and uu2,uu3,u≁xu\sim u_{2},u\sim u_{3},u\not\sim x. The following case shows that if uu exists, then χ(G)7\chi(G)\leq 7. In other words, if GG contains codomino3co-domino3, then χ(G)7\chi(G)\leq 7.

Partition A2=(N(v1)N(v2))(N(v2)N(v3))(N(v1)N(v3))A_{2}=(N(v_{1})\cap N(v_{2}))\cup(N(v_{2})\cap N(v_{3}))\cup(N(v_{1})\cup N(v_{3})) into following five vertex sets.

D1:=N(v1)N(v2)N(u2)N(u3)D_{1}:=N(v_{1})\cap N(v_{2})\cap N(u_{2})\cap N(u_{3}).
D2:=N(v1)N(v2)N(u1)N(u3)D_{2}:=N(v_{1})\cap N(v_{2})\cap N(u_{1})\cap N(u_{3}).
D3:=N(v2)N(v3)N(u1)N(u2)D_{3}:=N(v_{2})\cap N(v_{3})\cap N(u_{1})\cap N(u_{2}).
D4:=N(v2)N(v3)N(u1)N(u3).D_{4}:=N(v_{2})\cap N(v_{3})\cap N(u_{1})\cap N(u_{3}).
D5:=N(v1)N(v3).D_{5}:=N(v_{1})\cap N(v_{3}).

case2: Suppose GG contains codomino3co-domino3.

Suppose {u,u}E[D1,D3]\{u,u^{\prime}\}\subseteq E[D_{1},D_{3}].

χ(B1B3{u1,u3})3\chi(B_{1}\cup B_{3}-\{u_{1},u_{3}\})\leq 3. Define I:=B1B3{u1,u2,u3}I:=B_{1}\cup B_{3}-\{u_{1},u_{2},u_{3}\}. Recalling that B1B3A0B_{1}\cup B_{3}-A_{0} is anticomplete to {u2}\{u_{2}\} and that A0A_{0} is anticomplete to {v1,v2,v3}\{v_{1},v_{2},v_{3}\}, II is anticomplete to {v2,u2}\{v_{2},u_{2}\}. Therefore, IN(u)I-N(u) is anticomplete to {v2,u,u2}\{v_{2},u,u_{2}\}. Since G[{v2,u,u2}]P3G[\{v_{2},u,u_{2}\}]\simeq P_{3}, IN(u)I-N(u) is a stable set and hence χ(IN(u))1\chi(I-N(u))\leq 1. Similarly, χ((IN(u))N(u))1\chi((I\cap N(u))-N(u^{\prime}))\leq 1. Trivially, (IN(u)N(u)){u2}(I\cap N(u)\cap N(u^{\prime}))\cup\{u_{2}\} is stable set and hence the chromatic number is less than 2. Therefore, χ(G)χ(IN(u))+χ((IN(u))N(u))+χ((IN(u)N(u)){u2})1+1+1=3.\chi(G)\leq\chi(I-N(u))+\chi((I\cap N(u))-N(u^{\prime}))+\chi((I\cap N(u)\cap N(u^{\prime}))\cup\{u_{2}\})\leq 1+1+1=3.

G[D1{u1}{x}{v3}]G[D_{1}\cup\{u_{1}\}\cup\{x\}\cup\{v_{3}\}] is a stable set. If E[{x},D1]E[\{x\},D_{1}]\neq\emptyset, then GG contains codomino2co-domino2. Accoding to definition, E[{u1},D1{x}]=E[\{u_{1}\},D_{1}\cup\{x\}]=\emptyset. The rest is obvious.

G[D2D4(B2{x})]G[D_{2}\cup D_{4}\cup(B_{2}-\{x\})] is a stable set. Recalling that |C3|1|C_{3}|\leq 1 and C1=C2=C_{1}=C_{2}=\emptyset, B2{x}=C4B_{2}-\{x\}=C_{4}. According to definition D2D4(B2{x})D_{2}\cup D_{4}\cup(B_{2}-\{x\}) is complete to {u1,u3}\{u_{1},u_{3}\} and the rest is obvious.

G[D3{u3}{v1}]G[D_{3}\cup\{u_{3}\}\cup\{v_{1}\}] is a stable set. According to definition, E[D3,u3]=E[D_{3},u_{3}]=\emptyset and G[E3]G[E_{3}] is a stable set. The rest is obvious.

Let D5D_{5}^{\prime} be D5{v2}D_{5}\cup\{v_{2}\}. χ(G)χ(B1B3{u1,u3})+χ(D1{u1}{x}{v3})+χ(D2D4(B2{x}))+χ(D3{u3}{v1})4+1+1+1=7.\chi(G)\leq\chi(B_{1}\cup B_{3}-\{u_{1},u_{3}\})+\chi(D_{1}\cup\{u_{1}\}\cup\{x\}\cup\{v_{3}\})+\chi(D_{2}\cup D_{4}\cup(B_{2}-\{x\}))+\chi(D_{3}\cup\{u_{3}\}\cup\{v_{1}\})\leq 4+1+1+1=7.

Suppose E[D1,D3]=E[D_{1},D_{3}]=\emptyset.

G[D1D3{x}]G[D_{1}\cup D_{3}\cup\{x\}] is a stable set. If E[{x},D1D3]E[\{x\},D_{1}\cup D_{3}]\neq\emptyset, then GG contains codomino2co-domino2. Since E[D1,D3]=E[D_{1},D_{3}]=\emptyset, G[D1D3{x}]G[D_{1}\cup D_{3}\cup\{x\}] .

χ(G)χ(D1D3{x})+χ(D2D4(B2{x}))+χ(B1{v2,v3})+χ(B3{v1})+χ(D5)1+1+2+2+17\chi(G)\leq\chi(D_{1}\cup D_{3}\cup\{x\})+\chi(D_{2}\cup D_{4}\cup(B_{2}-\{x\}))+\chi(B_{1}\cup\{v_{2},v_{3}\})+\chi(B_{3}\cup\{v_{1}\})+\chi(D_{5})\leq 1+1+2+2+1\leq 7.

Combining codomino2co-domino2-free, dodomino3do-domino3-free and our supposition that evrey vertex in A2N(v2)A_{2}\cap N(v_{2}) has two neighbors in {u1,u2,u3}\{u_{1},u_{2},u_{3}\}, if uN(v1)N(v2)u\in N(v_{1})\cap N(v_{2}), then uu is complete to {u1,u3}\{u_{1},u_{3}\}. Similarly, if uN(v2)N(v3)u\in N(v_{2})\cap N(v_{3}) ,then uu is complete to {u1,u3}\{u_{1},u_{3}\}. Therefore, G[(N(v1)N(v2))(N(v2)N(v3))]G[(N(v_{1})\cap N(v_{2}))\cup(N(v_{2})\cap N(v_{3}))] is a stable set.

Furthermore,χ(N(v2){v1,v3})2\chi(N(v_{2})-\{v_{1},v_{3}\})\leq 2. Noticing that N(v2)=(N(v1)N(v2))(N(v3)N(v2))C3C4N(v_{2})=(N(v_{1})\cap N(v_{2}))\cup(N(v_{3})\cap N(v_{2}))\cup C_{3}\cup C_{4}, the proof is easy.

χ(G)χ(N(v2){v1,v3})+χ(B1{v2,v3})+χ(B3{v1})+χ(N(v1)N(v3))2+2+2+1=7.\chi(G)\leq\chi(N(v_{2})-\{v_{1},v_{3}\})+\chi(B_{1}\cup\{v_{2},v_{3}\})+\chi(B_{3}\cup\{v_{1}\})+\chi(N(v_{1})\cap N(v_{3}))\leq 2+2+2+1=7.

Therefore, if C3C_{3}\neq\emptyset, χ(G)7.\chi(G)\leq 7.

Suppose C1=C2=C3=C_{1}=C_{2}=C_{3}=\emptyset. We continue to use notations D1,D2,,D5D_{1},D_{2},...,D_{5}. Similar to case2, χ(G)7\chi(G)\leq 7.(The proof does not depend on the existence on {x}.)∎

Let C=v1v2v3u3u2u1C=v_{1}v_{2}v_{3}u_{3}u_{2}u_{1} be a 66-hole and {u}\{u\} be a vertex outside the 66-hole. We use X1X_{1} to denote a graph obtained from {u}C\{u\}\cup C by connecting edges v1v3,uv1v_{1}v_{3},uv_{1} and uu2uu_{2}.

Let C=v1v2v3u3u2u1C=v_{1}v_{2}v_{3}u_{3}u_{2}u_{1} be a 66-hole and {u}\{u\} be a vertex outside the 66-hole. We use X2X_{2} to denote a graph obtained from {u}C\{u\}\cup C by connecting edges v1v3,v3u1v_{1}v_{3},v_{3}u_{1},u1u3,v2uu_{1}u_{3},v_{2}u and uu2uu_{2}.

The following figures are X1X_{1} and X2X_{2}( from left to right).

[Uncaptioned image]
lemma 3.5.

Suppose GG is codominoco-domino-free and contains X1X_{1} or X2X_{2}, then χ(G)7\chi(G)\leq 7.

Proof.

case1:G contains X1X_{1}.

G[B1A0{u1}]G[B_{1}-A_{0}-\{u_{1}\}] is anticomplete to {u1,u2}\{u_{1},u_{2}\} and complete to {u,u3}\{u,u_{3}\}.Suppose yB1A0y\in B_{1}-A_{0} and yu1y\sim u_{1}(or u2u_{2}) only, then G[{y,u1,u2}{v3,v1}]P3P2G[\{y,u_{1},u_{2}\}\cup\{v_{3},v_{1}\}]\simeq P_{3}\cup P_{2}. Therefore, yy must complete to both u2u_{2} and u1u_{1} and hence we get G[{y,u1,u2}{v1,v3}]K3P2G[\{y,u_{1},u_{2}\}\cup\{v_{1},v_{3}\}]\simeq K_{3}\cup P_{2},which contradics our assumption. G[B2]G[B_{2}] is complete to {u3}\{u_{3}\}(or uu) or G[(B2A0){v1,v2}{u3,u2}]G[(B_{2}-A_{0})\cup\{v_{1},v_{2}\}\cup\{u_{3},u_{2}\}] will induce an P3P2P_{3}\cup P_{2}(G[(B2A0){v1,v2}{u,u2}]G[(B_{2}-A_{0})\cup\{v_{1},v_{2}\}\cup\{u,u_{2}\}] will induce a P3P2P_{3}\cup P_{2}).

Symmetrically , G[B3{u3}]G[B_{3}-\{u_{3}\}] is anticomplete to {u2,u3}\{u_{2},u_{3}\} and complete to {u,u1}\{u,u_{1}\} and G[B2]{u}G[B_{2}]-\{u\} is anticomplete to {u1,u3}\{u_{1},u_{3}\} and anticomplete to {u,u2}\{u,u_{2}\}.

Either G[B1A0]G[B_{1}-A_{0}] or G[B3]G[B_{3}] is edge-free. According to the above paragraph, G[(B1A0)B3]G[(B_{1}-A_{0})\cup B_{3}] is complete to uu and hence it is K3K_{3}-free. Suppose there is an edge in G[B1A0]G[B_{1}-A_{0}](y1,y2y_{1},y_{2}) and an edge in G[B3]G[B_{3}](y3,y4y_{3},y_{4}).

We prove G[v4,y1,y2,y3,y4,v6]codominoG[v_{4},y_{1},y_{2},y_{3},y_{4},v_{6}]\simeq co-domino to obtain a contradicion. It is easy to see that y3y_{3} should have at least one neighbor in {y2,y1}\{y_{2},y_{1}\}. So does y4y_{4}. Consequently, G[{y1,y2,y3,y4}]C4G[\{y_{1},y_{2},y_{3},y_{4}\}]\simeq C_{4}, otherwise G[{x1,x2,y1,y2}]G[\{x_{1},x_{2},y_{1},y_{2}\}] must contain K3K_{3}.

Symmetrically, either G[B1A0]G[B_{1}-A_{0}] or G[B2]G[B_{2}] is edge-free and either G[B2]G[B_{2}] or G[B3]G[B_{3}] is edge-free.

χ(B1B2B3)4\chi(B_{1}\cup B_{2}\cup B_{3})\leq 4. Recalling that G[Bi]G[B_{i}] is P3P_{3}-free. Suppose G[B1A0]G[B_{1}-A_{0}] has edge, then χ(B2)1,χ(B3)1\chi(B_{2})\leq 1,\chi(B_{3})\leq 1. If G[B2]G[B_{2}] or G[B3]G[B_{3}] contains edge, then we can prove χ(B1B2B3)4\chi(B_{1}\cup B_{2}\cup B_{3})\leq 4. If G[B1A0]G[B_{1}-A_{0}],G[B2]G[B_{2}] and G[B3]G[B_{3}] are all edge-free, then it is trvial that χ(B1B2B3)4\chi(B_{1}\cup B_{2}\cup B_{3})\leq 4.

Therefore, χ(G)χ(B1B2B3)+χ(A2)4+3=7.\chi(G)\leq\chi(B_{1}\cup B_{2}\cup B_{3})+\chi(A_{2})\leq 4+3=7.

case2: GG is X1X_{1}-free and contains X2X_{2}.

Accoring to section 2.1, we partition V(G)V(G) around G[{v1,v2,v3}]G[\{v_{1},v_{2},v_{3}\}].

Every vertex in B2{u}B_{2}-\{u\} is adjacent to u3u_{3} and not adjacent to u2u_{2}. If there is xB2{u}x\in B_{2}-\{u\} such that xx is not adjacent to u3u_{3}, then xx is adjacent to u2u_{2}. Otherwise G[{v1,v2,x}{u2,u3}]P3P2G[\{v_{1},v_{2},x\}\cup\{u_{2},u_{3}\}]\simeq P_{3}\cup P_{2} or K3P2K_{3}\cup P_{2}. Therefore, we can simply suppose there is xB2{u}x\in B_{2}-\{u\} such that xx is adjacent to u2u_{2}. However, G[{u,u2,x}{v1,v3}]P3P2G[\{u,u_{2},x\}\cup\{v_{1},v_{3}\}]\simeq P_{3}\cup P_{2} or K3P2K_{3}\cup P_{2}(depending on the adjacency of uu and xx). Therefore, such xx does not exist.

Every vertex in B1A0B_{1}-A_{0} is adjacent to {u3}\{u_{3}\}.

Every vertex in B3B_{3} is anticomplete to {u2,u3}\{u_{2},u_{3}\}.

Partition B1B2B3A0B_{1}\cup B_{2}\cup B_{3}-A_{0} into the folloing six sets.

  • D1:={x(B2{u})|x is adjacent to u1 and u3}D_{1}:=\{x\in(B_{2}-\{u\})|x\text{ is adjacent to }u_{1}\text{ and }u_{3}\}.

  • D2:={x(B2{u})|x is adjacent to u3 only}D_{2}:=\{x\in(B_{2}-\{u\})|x\text{ is adjacent to }u_{3}\text{ only}\}.

  • D3:={x(B1A0)|x is adjacent to u3 only}D_{3}:=\{x\in(B_{1}-A_{0})|x\text{ is adjacent to }u_{3}\text{ only}\}.

  • D4:={x(B1A0)|x is adjacent to u1 and u3}.D_{4}:=\{x\in(B_{1}-A_{0})|x\text{ is adjacent to }u_{1}\text{ and }u_{3}\}.

  • D5:={xB3|x is adjacent to u1}.D_{5}:=\{x\in B_{3}|x\text{ is adjacent to }u_{1}\}.

  • D6:={xB3|x has no neighbor in {u1,u2,u3}}.D_{6}:=\{x\in B_{3}|x\text{ has no neighbor in }\{u_{1},u_{2},u_{3}\}\}.

Since D1D4D_{1}\cup D_{4} is compete to {u1,u3}\{u_{1},u_{3}\} and u1u3u_{1}\sim u_{3}, G[D1D4]G[D_{1}\cup D_{4}] has no edge and hence χ(D1D4)1\chi(D_{1}\cup D_{4})\leq 1.

Since D2D_{2} and D3D_{3} anticomplete to {v3,u1,u2}\{v_{3},u_{1},u_{2}\}, G[D2D3]G[D_{2}\cup D_{3}] is P2P_{2}-free and hence χ(D2D3)1\chi(D_{2}\cup D_{3})\leq 1.

E[{u},A0D6]=E[\{u\},A_{0}\cup D_{6}]=\emptyset. Since {u}D6\{u\}\cup D_{6} is anticomplete to {v1,u1,u3}\{v_{1},u_{1},u_{3}\} and G[{v1,u1,u3}]P3G[\{v_{1},u_{1},u_{3}\}]\simeq P_{3}, G[{u}D6]G[\{u\}\cup D_{6}] has no edge. Since {u}A0\{u\}\cup A_{0} is anticomplete v1,v3,u3v_{1},v_{3},u_{3} and G[{v1,v3,u3}]P3G[\{v_{1},v_{3},u_{3}\}]\simeq P_{3}(It is trivial that u3u_{3} is anticomplete to A0A_{0}), G[{u}A0]G[\{u\}\cup A_{0}] has no edge.

Since D6A0D_{6}\cup A_{0} is anticomplete to {v2,v1,u1}\{v_{2},v_{1},u_{1}\} and G[{v2,v1,u1}]P3G[\{v_{2},v_{1},u_{1}\}]\simeq P_{3}, G[A0D6]G[A_{0}\cup D_{6}] has no edge and hence G[{u}D6A0]G[\{u\}\cup D_{6}\cup A_{0}] has no edge.

χ(B1B2B3)χ(D1D4)+χ(D2D3)+χ(A0{u}D6)+χ(D5)4\chi(B_{1}\cup B_{2}\cup B_{3})\leq\chi(D_{1}\cup D_{4})+\chi(D_{2}\cup D_{3})+\chi(A_{0}\cup\{u\}\cup D_{6})+\chi(D_{5})\leq 4. Therefore, χ(G)χ(B1B2B3)+χ(A2)7.\chi(G)\leq\chi(B_{1}\cup B_{2}\cup B_{3})+\chi(A_{2})\leq 7.

Suppose all graphs are (codomino,K3P2,X1,X2)(co-domino,K_{3}\cup P_{2},X_{1},X_{2})-free.

Let C=v1v2v3v4v5C=v_{1}v_{2}v_{3}v_{4}v_{5} be a 55-hole. We use cotwinC5co-twin-C_{5} to denote a graph obtained from CC by adding a vertex {v6}\{v_{6}\} which is only adjacent {v3,v4,v5}\{v_{3},v_{4},v_{5}\}.

[Uncaptioned image]

Let C=v1v2v3u3u2u1C=v_{1}v_{2}v_{3}u_{3}u_{2}u_{1} be a 66-hole and {u}\{u\} be a vertex outside the 66-hole. We use 𝒴\mathcal{Y} to denote a family of graphs obtained from {u}C\{u\}\cup C by connecting edge uv1,uv2uv_{1},uv_{2} and uu2uu_{2} and it does not matter whether uu1uu_{1} is connected or not.

[Uncaptioned image]

Define M(v1,v2):=GN(v1)N(v2){v1,v2}M(v_{1},v_{2}):=G-N(v_{1})-N(v_{2})-\{v_{1},v_{2}\}. In this lemma, N(v1)N(v_{1}) does not include {v2}\{v_{2}\} and N(v2)N(v_{2}) does not include {v1}\{v_{1}\}. We introduce a partition which is useful in the following lemma.

observation 3.

If GG contains cotwinC5co-twin-C_{5}. V(G)V(G) can be partitioned into {v1,v2},N(v2)N(v1)\{v_{1},v_{2}\},N(v_{2})\cup N(v_{1}) and M(v1,v2)M(v_{1},v_{2}).

lemma 3.6.

If GG contains cotwinC5co-twin-C_{5}, then χ(G)7\chi(G)\leq 7.

Proof.

case1: G contains an element of 𝒴\mathcal{Y}.

G[B1A0]G[B_{1}-A_{0}] is complete to {u3}\{u_{3}\} and anticomplete to {u1,u2}\{u_{1},u_{2}\}. Otherwise, suppose there is xG[B1A0]x\in G[B_{1}-A_{0}] which is not adjacent to {u3}\{u_{3}\} or has neighbor in {u1,u2}\{u_{1},u_{2}\}. If xx has neighbor in {u1,u2}\{u_{1},u_{2}\}, then G[{x,u1,u2}{v2,v3}]G[\{x,u_{1},u_{2}\}\cup\{v_{2},v_{3}\}] induces P3P2P_{3}\cup P_{2} or K3P2K_{3}\cup P_{2}. Therefore, xx is not adjacent to u3u_{3} and hence G[{v2,x}]P2G[\{v_{2},x\}]\simeq P_{2} is anticomplete to G[{u1,u2,u3}]P3G[\{u_{1},u_{2},u_{3}\}]\simeq P_{3}. Therefore, such xx does not exist.

Similarly, G[B3]G[B_{3}] is complete to {u1}\{u_{1}\} and anticomplete to {u3,u2}\{u_{3},u_{2}\}.

Either G[B1A0]G[B_{1}-A_{0}] or G[B3]G[B_{3}] has no edge. Otherwise, suppose there is an edge(y1,y2y_{1},y_{2}) in G[B1A0]G[B_{1}-A_{0}] and an edge(y3,y4y_{3},y_{4}) in G[B3]G[B_{3}]. If G[{y1,y2,y3,y4}]G[\{y_{1},y_{2},y_{3},y_{4}\}] induces diamonddiamond or C4C_{4}, then G[{u1,u2,u3,y1,y2,y3,y4}]G[\{u_{1},u_{2},u_{3},y_{1},y_{2},y_{3},y_{4}\}] induces X1X_{1} or X2X_{2}. Therefore, G[{y1,y2,y3,y4}]G[\{y_{1},y_{2},y_{3},y_{4}\}] is not isomorphic to diamonddiamond, C4C_{4} and K4K_{4}. However, such condition requires G[{y1,y2,y3,y4}]G[\{y_{1},y_{2},y_{3},y_{4}\}] to be isomorphic to P4P_{4} or 2K22K_{2}. Therefore, there must be a vertex has degree one in G[{y1,y2,y3,y4}]G[\{y_{1},y_{2},y_{3},y_{4}\}]. Suppose d(y1)=1d(y_{1})=1 in G[{y1,y2,y3,y4}G[\{y_{1},y_{2},y_{3},y_{4}\}. Then y1y_{1} is anticomplete to {y3,y4}\{y_{3},y_{4}\} and hence G[{u2,u3,y1}{y3,y4}]P3P2G[\{u_{2},u_{3},y_{1}\}\cup\{y_{3},y_{4}\}]\simeq P_{3}\cup P_{2}. Therefore, either y1,y2y_{1},y_{2} does not exist or {y3,y4}\{y_{3},y_{4}\} does not exists.

G[B2]G[B_{2}] is complete to {u3}\{u_{3}\}. Otherwise, suppose there exists zG[B2]z\in G[B_{2}] such that z is not adjacent to {u3}\{u_{3}\}. zz is adjacent to u2u_{2}, otherwise G[{v1,v2,z}{u2,u3}]P3P2G[\{v_{1},v_{2},z\}\cup\{u_{2},u_{3}\}]\simeq P_{3}\cup P_{2}. Since G[{x,u1,u2,u3,v1,v2,v3}]X1G[\{x,u_{1},u_{2},u_{3},v_{1},v_{2},v_{3}\}]\simeq X_{1}, zz is adjacent to u1u_{1}. However, if zz is adjacent to u1u_{1}, then G[{z,u1,u2,v1,v2,v3}]codominoG[\{z,u_{1},u_{2},v_{1},v_{2},v_{3}\}]\simeq co-domino. Therefore, such zz does not exist.

G[B2]G[B_{2}] contains at most one edge. Suppose there are two edges in G[B3]G[B_{3}]({y1,y2}\{y_{1},y_{2}\} and {y3,y4}\{y_{3},y_{4}\} ). Because v2v_{2} is complete to {u,y1,y2,y3,y4}\{u,y_{1},y_{2},y_{3},y_{4}\}. uu has at most one neighbor in {y1,y2}\{y_{1},y_{2}\} and {y3,y4}\{y_{3},y_{4}\}. Suppose u≁y2,u≁y3u\not\sim y_{2},u\not\sim y_{3}. Since both y2,y3y_{2},y_{3} are not adjacent to v1v_{1}, G[{y3,u3,y2}{u,v1}]P3P2G[\{y_{3},u_{3},y_{2}\}\cup\{u,v_{1}\}]\simeq P_{3}\cup P_{2}, which causes contradiction. Therefore, G[B2]G[B_{2}] has at most one edge.

Suppose B1A0B_{1}-A_{0} has no edge. Let B2=V1,B3=V2,A0=V3B_{2}=V_{1},B_{3}=V_{2},A_{0}=V_{3}, then we can apply claim3.3 to get χ(B1B2B3)χ(B1A0)+χ(B2B3A04\chi(B_{1}\cup B_{2}\cup B_{3})\leq\chi(B_{1}-A_{0})+\chi(B_{2}\cup B_{3}\cup A_{0}\leq 4 and hence χ(G)χ(B1B2B3)4+3=7\chi(G)\leq\chi(B_{1}\cup B_{2}\cup B_{3})\leq 4+3=7.

case2: G is 𝒴\mathcal{Y}-free and contains a cotiwnC5co-tiwn-C_{5}.

According to observation 3, V(G)={v1,v2}(N(v2)N(v1))M(v1,v2)V(G)=\{v_{1},v_{2}\}\cup(N(v_{2})\cup N(v_{1}))\cup M(v_{1},v_{2}).

Noticing that vertices in N(v1)N(v2){v3,v5}N(v_{1})\cup N(v_{2})-\{v_{3},v_{5}\} have neighbor in {v4,v6}\{v_{4},v_{6}\}, we can divide N(v1)N(v2){v3,v5}N(v_{1})\cup N(v_{2})-\{v_{3},v_{5}\} to make the structure more clearly. We difine two disjoint sets as following:

  • D1:={y|yN(v1)N(v2){v3,v5},yD_{1}:=\{y|y\in N(v_{1})\cup N(v_{2})-\{v_{3},v_{5}\},y is adjacent to only one vertex in {v3,v4,v5,v6}}\{v_{3},v_{4},v_{5},v_{6}\}\}

  • D2:={y|yN(v1)N(v2){v3,v5},yD_{2}:=\{y|y\in N(v_{1})\cup N(v_{2})-\{v_{3},v_{5}\},y is adjacent to more than one vertices in {v3,v4,v5,v6}}\{v_{3},v_{4},v_{5},v_{6}\}\}

|D1|=|D_{1}|=\emptyset. Otherwise, suppose y1D1y_{1}\in D_{1} and y1v1y_{1}\sim v_{1}. Suppose y1y_{1} is adjacent to {v4}\{v_{4}\}. Because if the only neighbor of y1y_{1} belongs to {v3,v5}\{v_{3},v_{5}\} , then G[{v1,v2,y1,v4,v6}]G[\{v_{1},v_{2},y_{1},v_{4},v_{6}\}] induces K3P2K_{3}\cup P_{2} or P3P2P_{3}\cup P_{2}.

If y1y_{1} is not adjacent to v2v_{2}, then G[{v1,y1,v2,v3,v4,v5,v6}]G[\{v_{1},y_{1},v_{2},v_{3},v_{4},v_{5},v_{6}\}] is isomorphic to an element in 𝒴\mathcal{Y}(when u1u_{1} is not adjacent to uu. However, if y1y_{1} is not adjacent to v2v_{2}, then G[{v1,y1,v2,v4,v5,v6}]codominoG[\{v_{1},y_{1},v_{2},v_{4},v_{5},v_{6}\}]\simeq co-domino. Therefore, such y1y_{1} does not exist.

Any vertex in D2D_{2} is complete to an edge in G[{v3,v4,v5,v6}]G[\{v_{3},v_{4},v_{5},v_{6}\}]. Otherwise, there is zD2z\in D_{2} which is not complete to any edge in G[{v3,v4,v5,v6}]G[\{v_{3},v_{4},v_{5},v_{6}\}]. zz must be adjacent to {v3,v5}\{v_{3},v_{5}\}, which leads to the contradiction that G[{v1,v2,z,v4,v6}]G[\{v_{1},v_{2},z,v_{4},v_{6}\}] induces K3P2K_{3}\cup P_{2} or P2P3P_{2}\cup P_{3}.

{v1,v2}M(v1,v2)\{v_{1},v_{2}\}\cup M(v_{1},v_{2}) can be colored with no more than 22 colors.

Recalling that points in D2D_{2} are at least adjacent to one edge in G[{v3,v4,v5,v6}]G[\{v_{3},v_{4},v_{5},v_{6}\}], it is trivial that χ(D2)5\chi(D_{2})\leq 5. In summary, χ(G)χ({v1,v2}M(v1,v2))+χ(D2{v3,v4,v5,v6})2+57\chi(G)\leq\chi(\{v_{1},v_{2}\}\cup M(v_{1},v_{2}))+\chi(D_{2}\cup\{v_{3},v_{4},v_{5},v_{6}\})\leq 2+5\leq 7. ∎

Suppose all graphs are (codomino,K3P2,X1,X2,cotwinC5)(co-domino,K_{3}\cup P_{2},X_{1},X_{2},co-twin-C_{5})-free.

Let C=u1v1v2v3u3u2C=u_{1}v_{1}v_{2}v_{3}u_{3}u_{2} be a 66-hole. We use χ37\chi 37 to denote a graph obtained from CC by connecting edge v1v3v1v3.

[Uncaptioned image]
lemma 3.7.

If GG contains χ37\chi 37 or coco-AA, then χ(G)7\chi(G)\leq 7.

Proof.

case1: GG contains χ37\chi 37.

We prove G[B2]G[B_{2}] is edge-free. We obtain contradiction by proving G[{v1,v3,u1,u3,x1,x2}]cotwinC5G[\{v_{1},v_{3},u_{1},u_{3},x_{1},x_{2}\}]\simeq co-twin-C_{5}. In other words, contradiction comes out if x1,x2x_{1},x_{2} are anticomplete to {u2}\{u_{2}\} and complete to {u1,u3}\{u_{1},u_{3}\}. Suppose there is an edge in G[B2]G[B_{2}] and we call them G[x1,x2]G[x_{1},x_{2}]. If x1x_{1}(or x2x_{2}), G[{v1,v3}{x1,x2,u2}]G[\{v_{1},v_{3}\}\cup\{x_{1},x_{2},u_{2}\}] would induce K3P2K_{3}\cup P_{2} or P3P2P_{3}\cup P_{2}. Furthermore, we can prove {x1,x2}\{x_{1},x_{2}\} is complete to {u1,u3}\{u_{1},u_{3}\} without much difficulty and hence we finish our proof of proposion that G[B2]G[B_{2}] is edge-free.

Furthermore, either G[B3]G[B_{3}] or G[B1A0]G[B_{1}-A_{0}] has no edge. Otherwise, suppose there is {y1,y2}E(B1A0)\{y_{1},y_{2}\}\in E(B_{1}-A_{0}), {y3,y4}E(B3)\{y_{3},y_{4}\}\in E(B_{3}). Since B1A0B_{1}-A_{0} is complete to {u3}\{u_{3}\} and anticomplete to {u1,u2}\{u_{1},u_{2}\} and B3B_{3} is complete to {u1}\{u_{1}\} and anticomplete to {u2,u3}\{u_{2},u_{3}\}, G[y1,y3,u1,u3,u2,y2]G[y_{1},y_{3},u_{1},u_{3},u_{2},y_{2}] induces codominoco-domino or cotwinC5co-twin-C_{5}.

According to claim3.3, χ(B1B3)3\chi(B_{1}\cup B_{3})\leq 3. Therefore, χ(G)χ(B2)+χ(B1B3)+χ(A2)7\chi(G)\leq\chi(B_{2})+\chi(B_{1}\cup B_{3})+\chi(A_{2})\leq 7.

case2: GG is χ37\chi 37-free and contains a coco-AA.

G[B2]G[B_{2}] is complete to {v4,v6}\{v_{4},v_{6}\} and hence G[B2]G[B_{2}] is edge-free. Otherwise, there is xB2x\in B_{2} such that xx has only one neighbor in {v4,v6}\{v_{4},v_{6}\}. If xv4x\sim v_{4}, then xv5x\sim v_{5}. Otherwise, G[{v1,v2,x}{v5,v6}]G[\{v_{1},v_{2},x\}\cup\{v_{5},v_{6}\}] induces P3P2P_{3}\cup P_{2}. However, G[{x,v2,v3,v6,v5,v1}]G[\{x,v_{2},v_{3},v_{6},v_{5},v_{1}\}] induces χ37\chi 37. Therefore, xv6x\sim v_{6}. Now v≁v5v\not\sim v_{5}, otherwise G[{v1,v2,v3,x,v6,v5}]G[\{v_{1},v_{2},v_{3},x,v_{6},v_{5}\}] induces codominoco-domino. However, G[{v1,v2,x,v6,v4,v5}]G[\{v_{1},v_{2},x,v_{6},v_{4},v_{5}\}] induces coAco-A. Therefore, such xx does not exist.

every edge in G[B1A0]G[B_{1}-A_{0}] includes one vertex which is complete to {v4,v6}\{v_{4},v_{6}\}. Otherwise, there is an edge in G[B1A0]G[B_{1}-A_{0}] whose vertices has at most one neighbor in {v4,v6}\{v_{4},v_{6}\}. If one of them(xx) is adjacent to {v4}\{v_{4}\} rather than {v6}\{v_{6}\},then G[{v2,v1,x}{v5,v6}]G[\{v_{2},v_{1},x\}\cup\{v_{5},v_{6}\}] induces P3P2P_{3}\cup P_{2}. Therefore, none of these vertices is adjacent to {v4}\{v_{4}\} and hence G[{v3,v4,v5}B1A0]G[\{v_{3},v_{4},v_{5}\}\cup B_{1}-A_{0}] induces P3P2P_{3}\cup P_{2}. We already obtain a contradiction.

According to the above discussion, G[(B1A0)B2]G[(B_{1}-A_{0})\cup B_{2}] can be partitioned into two stable sets: the vertices which are complete to {v4,v6}\{v_{4},v_{6}\} and the rest of vertices in B1A0B_{1}-A_{0}. Therefore, χ((B1A0)B2)2\chi((B_{1}-A_{0})\cup B_{2})\leq 2.

Therefore, χ(G)χ(B2(B1A0))+χ(B3A0)+χ(A2)2+2+3=7.\chi(G)\leq\chi(B_{2}\cup(B_{1}-A_{0}))+\chi(B_{3}\cup A_{0})+\chi(A_{2})\leq 2+2+3=7.

Combining lemma3.4 and lemma3.7, we obtain theorem1.2 straightly.

theorem.

1.2 If GG contains codominoco-domino or coAco-A, then χ(G)7\chi(G)\leq 7.

Suppose GG is (K3P2,codomino,coA)(K_{3}\cup P_{2},co-domino,co-A)-free with G[D1]¯\overline{G[D_{1}]} is P3P_{3}-free.

claim 3.8.

If v1,v2D1v_{1},v_{2}\in D_{1} and v1≁v2v_{1}\not\sim v_{2}, then either G[N(v1)N(v2)]G[N(v_{1})-N(v_{2})] or G[N(v2)N(v1)]G[N(v_{2})-N(v_{1})] has no edge.

Proof.

Suppose both sets have edges. The vertex sets in G[N(v1)N(v2)]G[N(v_{1})-N(v_{2})] is {u1,u2}\{u_{1},u_{2}\} and that of edge in G[N(v2)N(v1)]G[N(v_{2})-N(v_{1})] is {u3,u4}\{u_{3},u_{4}\}. G[{v1,v2,u1,u2,u3,u4}]G[\{v_{1},v_{2},u_{1},u_{2},u_{3},u_{4}\}] must induce codominoco-domino or coAco-A or P3P2P_{3}\cup P_{2}. Therefore, either G[N(v1)N(v2)]G[N(v_{1})-N(v_{2})] or G[N(v2)N(v1)]G[N(v_{2})-N(v_{1})] is edge-free. ∎

theorem.

(1.3) If G[D2]G[D_{2}] is not a clique, then χ(G)\chi(G)\leq7.

Proof.

Since G[D2]G[D_{2}] is not a clique, G[D2]G[D_{2}] has two nonadjacent vertices {v1,v2}\{v_{1},v_{2}\}. According to claim3.8, either G[N(v1)N(v2)]G[N(v_{1})-N(v_{2})] or G[N(v2)N(v1)]G[N(v_{2})-N(v_{1})] has no edge. Without loss of generality, we suppose G[N(v1)N(v2)]G[N(v_{1})-N(v_{2})] is edge-free and hence χ(N(v1)N(v2))1\chi(N(v_{1})\cup N(v_{2}))\leq 1.

If ω(N(v1)N(v2))1\omega(N(v_{1})\cap N(v_{2}))\leq 1, then χ(N(v1)N(v2))1\chi(N(v_{1})-N(v_{2}))\leq 1 or χ(N(v2)N(v1))1\chi(N(v_{2})-N(v_{1}))\leq 1. We suppose χ(N(v1)N(v2))1\chi(N(v_{1})-N(v_{2}))\leq 1, then χ(v1)2\chi(v_{1})\leq 2. If G[N(v1)]G[N(v_{1})] has edge({v2,v3}\{v_{2},v_{3}\}), then accoring to section 2.1, G[{v1,v2,v3}]G[\{v_{1},v_{2},v_{3}\}] can be divided into B1B2B3(N(v1)N(v2))(N(v1)N(v3))(N(v2)N(v3))B_{1}\cup B_{2}\cup B_{3}\cup(N(v_{1})\cap N(v_{2}))\cup(N(v_{1})\cap N(v_{3}))\cup(N(v_{2})\cap N(v_{3})).

χ(N(v1))χ(N(v1)N(v2))+χ(N(v1)N(v2))2\chi(N(v_{1}))\leq\chi(N(v_{1})\cap N(v_{2}))+\chi(N(v_{1})-N(v_{2}))\leq 2. The vertex sets left to be colored are:A0,N(v2)N(v3),B2A_{0},N(v_{2})\cap N(v_{3}),B_{2} and B3B_{3}. Since χ(B2A0)2\chi(B_{2}\cup A_{0})\leq 2, χ(B3)2\chi(B_{3})\leq 2 and χ(N(v2)N(v3))1\chi(N(v_{2})\cap N(v_{3}))\leq 1. Therefore, χ(G)7.\chi(G)\leq 7.

Suppose ω(N(v1)N(v2))=2\omega(N(v_{1})\cap N(v_{2}))=2.

We partition V(G)V(G) into three parts: {v1,v2}\{v_{1},v_{2}\}, N(v1)N(v2)N(v_{1})\cup N(v_{2}) and G{v1,v2}N(v1)N(v2)G-\{v_{1},v_{2}\}-N(v_{1})-N(v_{2}). Define A:=G{v1,v2}N(v1)N(v2)A:=G-\{v_{1},v_{2}\}-N(v_{1})-N(v_{2}).

χ(A)3\chi(A)\leq 3. Suppose there is an edge(x1x2x_{1}x_{2}) in G[A]G[A]. Then AA can be divided into AN(x1)A-N(x_{1}) and (AN(x2))N(x1)(A-N(x_{2}))\cap N(x_{1}) and AN(x1)N(x2)A\cap N(x_{1})\cap N(x_{2}). Since AN(x1)A-N(x_{1}) is anticomplete to {v1,x1,v2}\{v_{1},x_{1},v_{2}\} and G[{v1,x1,v2}]P3G[\{v_{1},x_{1},v_{2}\}]\simeq P_{3}, G[AN(x1)]G[A-N(x_{1})] must be P2P_{2}-free and hence χ(AN(x1))1\chi(A-N(x_{1}))\leq 1. Similarly, χ((AN(x2))N(x1))1\chi((A-N(x_{2}))\cap N(x_{1}))\leq 1. Because x1x2x_{1}\sim x_{2} and ω(G)3\omega(G)\leq 3, χ(AN(x1)N(x2))1\chi(A\cap N(x_{1})\cap N(x_{2}))\leq 1. Therefore ,χ(A)3\chi(A)\leq 3.

χ(N(v2))3\chi(N(v_{2}))\leq 3. According to definition of D2D_{2}, there is a K3({u1,u2,u3})K_{3}(\{u_{1},u_{2},u_{3}\}) induced in G[GN(v2)]G[G-N(v_{2})]. Define Ei:={xN(v2)|x only adjacent to {ui}}E_{i}:=\{x\in N(v_{2})|x\text{ only adjacent to }\{u_{i}\}\} and Ei,j:={xN(v2)|x only adjacent to {ui} and {uj}}E_{i,j}:=\{x\in N(v_{2})|x\text{ only adjacent to }\{u_{i}\}\text{ and }\{u_{j}\}\}. Since GG is coco-AA-free, EiE_{i} is anticomplete to Ei,i+1E_{i,i+1}(mod33). Obviously, |Ei|1|E_{i}|\leq 1 for i=1,2,3i=1,2,3 and G[Ei,j]G[E_{i,j}] is edge-free. Therefore χ(N(v2))i=13χ(EiEi,i+1)3\chi(N(v_{2}))\leq\sum_{i=1}^{3}\chi(E_{i}\cup E_{i,i+1})\leq 3.

Because χ(N(v2))3\chi(N(v_{2}))\leq 3, χ(N(v2)N(v1))χ(N(v1)N(v2))+χ(N(v2))1+3=4\chi(N(v_{2})\cup N(v_{1}))\leq\chi(N(v_{1})-N(v_{2}))+\chi(N(v_{2}))\leq 1+3=4. Consequently, χ(G)χ(N(v1)N(v2))+χ({v1,v2}A)4+3\chi(G)\leq\chi(N(v_{1})\cup N(v_{2}))+\chi(\{v_{1},v_{2}\}\cup A)\leq 4+3=7. ∎

Therefore, combining theorem1.1, theorem1.2 and theorem1.3, we obtain our major theorem.

theorem.

If GG is (P3P2,K4)(P_{3}\cup P_{2},K_{4})-free, then χ(G)7\chi(G)\leq 7.

Finally, we prove a simple theorem using the above theorem.

theorem.

(1.4) If GG is (4K1,P3P2¯)(4K_{1},\overline{P_{3}\cup P_{2}})-free with order nn and clique number ω\omega, then n7ωn\leq 7\omega and χ(G)4ω\chi(G)\leq 4\omega.

Proof.

Suppose GG is (4K1,P3P2¯)(4K_{1},\overline{P_{3}\cup P_{2}})-free.

If G¯\overline{G} is connected, then we apply the above theorem to obtain that χ(G¯)7\chi(\overline{G})\leq 7. Therefore, V(G¯)V(\overline{G}) can be partitioned into 77 stable sets and hence V(G)V(G) can be partitioned into 77 cliques. Therefore, n7ωn\leq 7\omega.

If G¯\overline{G} is not connnected, then there is G1G_{1} such that G=G1+G2G=G_{1}+G_{2} and hence |V(G)|=|V(G1)|+|V(G2)||V(G)|=|V(G_{1})|+|V(G_{2})|. Such decomposition will last until each GiG_{i} satidfied Gi¯\overline{G_{i}} is connected. Suppose G=G1+GkG=G_{1}+...G_{k}. Since V|Gi|7ω(Gi)V|G_{i}|\leq 7\omega(G_{i}) and ω=j=1kω(Gi)\omega=\sum_{j=1}^{k}\omega(G_{i}), |V(G)|=n7ω|V(G)|=n\leq 7\omega.

It is trivial that the complement of bipartite graph is perfect graph. Therefore V(G)V(G) can be partitioned into 33 perfect graphs and one clique. Therefore, χ(G)4ω\chi(G)\leq 4\omega.∎

4 acknowledge

I try to go further on (P3P2,K4)(P_{3}\cup P_{2},K_{4})-free graphs. However, I fail and 7 is the best upper bound I obtain. I know this is not a well-written paper. If you are confused about some details in the proof, you can email me at [email protected] or [email protected].

References

  • [1] A. Bharathi,S. Choudum. Colouring of (P3P2)(P_{3}\cup P_{2})-free graphs.Graphs and Combinatorics 34(1):97–107,2018.
  • [2] L. Esperet, L. Lemoine, F. Maffray, and G. Morel. The chromatic number of (P5,K4)(P_{5},K_{4})-free
  • [3] A. Gyárfás. On ramsey covering numbers. Coll. Math. Soc. János Bolyai, 10:801–816, 1973. graphs. Discrete Math, 313:743–754, 2013.
  • [4] T. Karthick and S. Mishra. Coloring (P6P_{6}, diamond, K4K_{4})-free graphs.arXiv:1703.00606, 2017.
  • [5] Rui Li, Jinfeng Li and Di Wu. On the chromatic number of some (P3P2P_{3}\cup P_{2})-free graphs, arXiv: 2308.15248, 2023.
  • [6] G. Serge and S. Huang. (2P2,K4)(2P_{2},K_{4})-Free Graphs are 4-Colorable. SIAM Joural on Discrete Mathematics 33:1095–1120,2019.