On Sum--Connectivity Index of Bicyclic Graphs
Zhibin Du, Bo Zhou

TL;DR
This paper investigates the extremal sum--connectivity indices of bicyclic graphs with given vertices and matching number, identifying the graphs that maximize or minimize this index.
Contribution
It determines the minimum, maximum, and second extremal sum--connectivity indices of bicyclic graphs for specified parameters and characterizes the extremal graphs.
Findings
Identified graphs with minimum sum--connectivity index.
Identified graphs with maximum sum--connectivity index.
Characterized extremal graphs for given parameters.
Abstract
We determine the minimum sum--connectivity index of bicyclic graphs with vertices and matching number , where , the minimum and the second minimum, as well as the maximum and the second maximum sum--connectivity indices of bicyclic graphs with vertices. The extremal graphs are characterized.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsGraph theory and applications · Graph Labeling and Dimension Problems · Alzheimer's disease research and treatments
On Sum–Connectivity Index of Bicyclic Graphs
Zhibin Du, Bo Zhou111Corresponding author. E-mail address: [email protected]
Department of Mathematics, South China Normal University,
Guangzhou 510631, P. R. China
Abstract
We determine the minimum sum–connectivity index of bicyclic graphs with vertices and matching number , where , the minimum and the second minimum, as well as the maximum and the second maximum sum–connectivity indices of bicyclic graphs with vertices. The extremal graphs are characterized.
MSC 2000: 05C90; 05C35; 05C07
Keywords: Sum–connectivity index; Randić connectivity index; Matching number; Bicyclic graphs
1 Introduction
The Randić connectivity index [8] is one of the most successful molecular descriptors in structure–property and structure–activity relationships studies, e.g., [9, 10]. Its mathematical properties as well as those of its generalizations have been studied extensively as summarized in the books [6, 5]. Recently, a closely related variant of Randić connectivity index called the sum–connectivity index was proposed in [14].
Let be a simple graph with vertex set and edge set . For , denotes the degree of in . The Randić connectivity index (or product–connectivity index [14, 7]) of the graph is defined as [8]
[TABLE]
The sum–connectivity index of is defined as [14]
[TABLE]
It has been found that the sum–connectivity index and the Randić connectivity index correlate well among themselves and with -electronic energy of benzenoid hydrocarbons [14, 7]. Some mathematical properties of the sum–connectivity index have been established in [14, 2, 3]. Recall that an -vertex connected graph is known as a tree, a unicyclic graph and a bicyclic graph if it possesses , and edges, respectively. In [2], we obtained the minimum sum–connectivity indices of trees and unicyclic graphs respectively with given number of vertices and matching number, and determined the corresponding extremal graphs. The -vertex trees with the first a few minimum and maximum sum–connectivity indices were determined in [14], while the -vertex unicyclic graphs with the first a few minimum and maximum sum–connectivity indices were determined in [2] and [3], respectively. In this paper, we consider the sum–connectivity indices of bicyclic graphs.
A matching of the graph is a subset of such that no two edges in share a common vertex. A matching of is said to be maximum, if for any other matching of , . The matching number of is the number of edges of a maximum matching in .
If is a matching of a graph and vertex is incident with an edge of , then is said to be -saturated, and if every vertex of is -saturated, then is a perfect matching.
In this paper, we obtain the minimum sum–connectivity index in the set of bicyclic graphs with vertices and matching number , where . We also determine the minimum and the second minimum, as well as the maximum and the second maximum sum–connectivity indices in the set of bicyclic graphs with vertices. The extremal graphs are characterized.
Study on the Randić connectivity indices of bicyclic graphs may be found in [6, 15, 1, 12], and in particular, the minimum and the maximum Randić connectivity indices in the set of bicyclic graphs with vertices were determined in [12] and [1], respectively.
We note that some other graph invariants based on end–vertex degrees of edges in a graph have been studied recently, see, e.g., [4, 11, 13].
2 Preliminaries
For , let be the set of bicyclic graphs with vertices and matching number .
For , let be the graph obtained by identifying a vertex of two triangles, and attaching pendent vertices (vertices of degree one) and paths on two vertices to the common vertex of the two triangles, see Fig.
- Obviously, .
\left\}\begin{array}[]{c}\\ \\ \\ \end{array}\right.$$n-2m+1$$\left\{\begin{array}[]{c}\\ \\ \\ \end{array}\right.$$m-3Fig. 1. The graph .
Let ba a cycle on vertices. Let be the set of -vertex bicyclic graphs without pendent vertices, where . Let be the set of bicyclic graphs obtained by joining two vertex–disjoint cycles and with by an edge, where . Let be the set of bicyclic graphs obtained by joining two vertex–disjoint cycles and with by a path of length , where . Let be the set of bicyclic graphs obtained by identifying a vertex of and a vertex of with , where . Let be the set of bicyclic graphs obtained from by adding an edge, where . Let be the set of bicyclic graphs obtained by joining two non-adjacent vertices of with by a path of length , where . Obviously, .
Let be the set of bicyclic graphs on vertices.
3 Minimum sum–connectivity index of bicyclic graphs with given matching number
First we give some lemmas that will be used.
For a graph with , denotes the graph resulting from by deleting the vertex (and its incident edges).
Lemma 3.1
[2]** Let be an -vertex connected graph with a pendent vertex , where . Let be the unique neighbor of , and let be a neighbor of different from .
- (i)
If and there is at most one pendent neighbor of in , then
[TABLE]
with equality if and only if one neighbor of has degree one, and the other neighbors of are of degree two. 2. (ii)
If there are at most pendent neighbors of in , then
[TABLE]
with equality if and only if neighbors of have degree one, and the other neighbors of are of degree two.
Lemma 3.2
[2]** (i) The function is decreasing for .
(ii) For integer , the function is decreasing for .
Lemma 3.3
[2]** Let be a connected graph with , where , , and and have no common neighbor in . Let be the graph obtained from by deleting the edge , identifying and , which is denoted by , and attaching a pendent vertex to . Then .
Lemma 3.4
For , , and for , .
Proof. Let for , and let for . Note that . Then , implying that , and , implying that .
Lemma 3.5
For ,
[TABLE]
with equality if and only if .
Proof. Let for . Then , implying that is increasing on . It is easily seen that
[TABLE]
with equality if and only if .
Let be the graph obtained by attaching a pendent vertex to every vertex of a triangle. For , let be the unicyclic graph obtained by attaching pendent vertices and paths on two vertices to one vertex of a triangle.
Lemma 3.6
[2]** Let be a unicyclic graph with vertices and perfect matching, where . Suppose that . Then
[TABLE]
with equality if and only if .
For an edge of the graph (the complement of , respectively), (, respectively) denotes the graph resulting from by deleting (adding, respectively) the edge .
Lemma 3.7
Let and no pendent vertex has neighbor of degree two, where . Then with equality if and only if and .
Proof. Let .
Since and no pendent vertex has neighbor of degree two, is obtainable by attaching some pendent vertices to a graph in , where , and any two pendent vertices have no common neighbor (if , then no pendent vertex is attached).
Case 1. There is no vertex of degree two in . Then either , is obtainable by attaching a pendent vertex to every vertex of a graph in , or , is obtainable by attaching a pendent vertex to every vertex with degree two of a graph in . By direct calculation, we find that for , for , and for . Thus by Lemma 3.4, we have .
Case 2. There is a vertex, say , of degree two in . Denote by and the two neighbors of in . Then one of the two edges incident with , say , where is a perfect matching of . Suppose that there is no vertex of degree two in any cycle of . Since no pendent vertex has neighbor of degree two in , lies on the path joining the two disjoint cycles of . For , the difference of the numbers of vertices of degree two outside any cycle of and is equal to one, and thus by Lemma 3.3, . Repeating the operation from to , we finally get a graph , which has no vertex of degree two, such that , and thus the result follows from Case 1. Now suppose that lies on some cycle of . Consider , which is a unicyclic graph with perfect matching. If , then is obtained from by adding an edge either between two pendent vertices, and thus , or between two neighbors of a vertex of degree three, one of which being a pendent vertex, and thus . In either case, . Suppose that . Then by Lemma 3.6, . Note that and has at most one pendent neighbor. By Lemmas 3.2 (i) and 3.5, we have
[TABLE]
with equalities if and only if , , and , i.e., .
By combining Cases 1 and 2, the result follows.
Lemma 3.8
Let . Then with equality if and only if .
Proof. If has a pendent vertex whose neighbor is of degree two, then is the graph obtained from the unique -vertex bicyclic graph by attaching a path on two vertices to either a vertex of degree three, or a vertex of degree two, and thus it is easily seen that . Otherwise, by Lemma 3.7, is the unique graph with the minimum sum–connectivity index.
Now we consider the bicyclic graphs with perfect matching. There is a unique bicyclic graph with four vertices, and its matching number is two.
Theorem 3.1
Let , where . Then
[TABLE]
with equality if and only if .
Proof. Let . We prove the result by induction on . If , then the result follows from Lemma 3.8.
Suppose that and the result holds for graphs in . Let with a perfect matching .
If there is no pendent vertex with neighbor of degree two in , then by Lemma 3.7, . Suppose that has a pendent vertex whose neighbor is of degree two. Then and . Let be the neighbor of different from . Since , we have . Note that there is at most one pendent neighbor of in . Then by Lemma 3.1 (i), Lemma 3.2 (i) and the induction hypothesis,
[TABLE]
with equalities if and only if and , i.e., .
In the following we consider the sum–connectivity indices of graphs in the set of bicyclic graphs with vertices and matching number . We first consider the case .
Lemma 3.9
[15]** Let with , and has at least one pendent vertex. Then there is a maximum matching and a pendent vertex such that is not -saturated.
Theorem 3.2
Let , where . Then
[TABLE]
with equality if and only if .
Proof. Let . We prove the result by induction on . If , then the result follows from Theorem 3.1. Suppose that and the result holds for graphs in . Let .
Suppose that there is no pendent vertex in . Then and . It is easily seen that there are exactly three values for , and thus we have with . Let for . Then , and thus , implying that , i.e., . Then .
Suppose that there is at least one pendent vertex in . By Lemma 3.9, there is a maximum matching and a pendent vertex of such that is not -saturated. Then . Let be the unique neighbor of . Since is a maximum matching, contains one edge incident with . Note that there are edges of outside . Then , i.e., . Let be the number of pendent neighbors of in . Since at least pendent neighbors of are not -saturated, we have , i.e., . By Lemma 3.1 (ii) with , Lemma 3.2 (ii) and the induction hypothesis,
[TABLE]
with equalities if and only if , and , i.e., .
Now we consider the sum–connectivity indices of graphs bicyclic graphs matching number two. Let be the graph obtained by attaching and pendent vertices to the two vertices of degree three of the unique -vertex bicyclic graph, respectively, where , and .
Lemma 3.10
Among the graphs in with , and are respectively the unique graphs with the minimum and the second minimum sum–connectivity indices, which are equal to and , respectively.
Proof. Let . Then may be of three types:
(a) with . Suppose that . Let for . Then , implying that is decreasing for . It is easily seen that
[TABLE]
and thus, for . It follows that and are respectively the unique graphs with the minimum and the second minimum sum–connectivity indices, which are equal to and , respectively.
(b) is the graph obtained by attaching pendent vertices to a vertex of degree two of the unique -vertex bicyclic graph. Then
[TABLE]
since , where is decreasing for .
(c) is the graph obtained by attaching some pendent vertices to one or two vertices of degree three of the unique -vertex bicyclic graph in , and by Lemma 3.3 and the arguments in case (a), .
Now the result follows easily.
4 Minimum sum–connectivity index of bicyclic graphs
In this section, we determine the minimum and the second minimum sum–connectivity indices of bicyclic graphs with vertices.
Theorem 4.1
Among the graphs in with , is the unique graph with the minimum sum–connectivity index, which is equal to , the graph obtained by attaching a pendent vertex to a vertex of degree two of the unique -vertex bicyclic graph for is the unique graph with the second minimum sum–connectivity index, which is equal to , for is the unique graph with the second minimum sum–connectivity index, which is equal to , and for is the unique graph with the second minimum sum–connectivity index, which is equal to .
Proof. There are five graphs in . Thus, the case may be checked directly. Suppose in the following that .
Let and the matching number of , where . If , then by Lemma 3.10, with equality if and only if . If , then by Theorem 3.2, with equality if and only if . If , then by Theorem 3.2 and Lemma 3.3, . Let for . Then , implying that is increasing for . Note that
[TABLE]
Thus is the unique graph with the minimum sum–connectivity index.
Suppose that . If , then by Lemma 3.10, with equality if and only if . By the arguments as above, the second minimum sum–connectivity index of graphs in is precisely achieved by the minimum one of and . If , then . Suppose that . Let for . Then , implying that is decreasing for . Note that
[TABLE]
and then . Thus for and for are the unique graphs with the second minimum sum–connectivity index among graphs in .
5 Maximum sum–connectivity index of bicyclic graphs
In this section, we determine the maximum and the second maximum sum–connectivity indices of bicyclic graphs with vertices. Let be the path on vertices.
Lemma 5.1
[14]** For a connected graph with at least two vertices and a vertex , let be the graph obtained from by attaching two paths and to , the graph obtained from by attaching a path to , where . Then .
Lemma 5.2
Suppose that is a connected graph with and . Let be the graph obtained from by attaching a path to . Denote by and the two neighbors of in , and the pendent vertex of the path attached to in . Let .
- (i)
If and the maximum degree of is at most five, then . 2. (ii)
If , and there are at least two neighbors of in with degree two and , then . 3. (iii)
If and all the neighbors of in are of degree two, then .
Proof. (i) If , then
[TABLE]
If , then
[TABLE]
(ii) There are two neighbors of with degree two, let be the degree of the third neighbor of in . If , then since , we have
[TABLE]
If , then since , we have
[TABLE]
(iii) If , then
[TABLE]
If , then
[TABLE]
The proof is now completed.
Let be the set of connected graphs on vertices with exactly two vertex–disjoint cycles. Let be the set of connected graphs on vertices with exactly two cycles of a common vertex. Let be the set of connected graphs on vertices with exactly two cycles with at least one edge in common. Obviously, . For , let be the distance between and in .
Lemma 5.3
Among the graphs in with , the graphs in and the graphs in are respectively the unique graphs with the maximum and the second maximum sum–connectivity indices, which are equal to and , respectively.
Proof. Suppose that is a graph in with the maximum sum–connectivity index, and and are its two cycles. Let and be the two vertices such that . Let be the path joining and . By Lemma 5.1, the vertices outside , and are of degree one or two, the vertices on , and different from and are of degree two or three, and or .
Suppose that . If there is some vertex, say , on , or different from and with degree three, then making use of Lemma 5.2 (i) to by setting , we may get a graph in with larger sum–connectivity index, a contradiction. Thus the vertices on , and different from and are of degree two. If , then making use of Lemma 5.2 (ii) to by setting , we may get a graph in with larger sum–connectivity index, a contradiction. Thus . Similarly, we have . It follows that .
Suppose that . Suppose that one of and , say , is of degree four. Then by Lemma 5.2 (i), the vertices on and different from and are of degree two. If , then making use of Lemma 5.2 (ii) to by setting , we may get a graph in with larger sum–connectivity index, a contradiction. Thus . Denote by the pendent vertex of the path attached to . Consider . If , then
[TABLE]
If , then
[TABLE]
In either case, with , a contradiction. Thus . Note that and by Lemma 5.2 (i), there is exactly one vertex, say , on and different from and with degree three. Denote by the pendent vertex of the path attached to . Consider . Let be the degree of the neighbor of , one neighbor of on is of degree two, and we denote by the other degree of the neighbor of on , where or . We have
[TABLE]
and thus, with , which is also a contradiction.
Now we have shown that the graphs in are the unique graphs in with the maximum sum–connectivity index. Note that for and ,
[TABLE]
The result follows.
Lemma 5.4
Among the graphs in with , the graphs in and the graphs in are respectively the unique graphs with the maximum and the second maximum sum–connectivity indices, which are equal to and , respectively.
Proof. Suppose that is a graph in with the maximum sum–connectivity index. Then has exactly three cycles, let and be its two cycles such that the remaining one is of the maximum length. Let be the set of the common vertices of and . Let and be the two vertices in such that . By Lemma 5.1, the vertices outside and are of degree one or two, the vertices on and different from and are of degree two or three, and or . Denote by (, respectively) the neighbor of on ( on , respectively) different from the vertices in .
If , then by Lemma 5.2 (i) and (ii), we have .
Suppose that . Suppose that the lengths of and are at least four. Consider . Note that
[TABLE]
If , then . If , then by Lemma 5.2 (i), we have , and thus . In either case, we have . By Lemma 5.3, we have for with equality if and only if , i.e., .
Suppose that at least one of and , say , is of length three. Since , there are some vertices outside and . By Lemma 5.2 (i) and (ii), the subgraph induced by the vertices outside and is a path, say , which is attached to . Suppose that . Denote by the neighbor of outside and . Consider . If or , then
[TABLE]
and thus , a contradiction. Hence , and the length of is at least four. Note that one neighbor of on is of degree two. Denote by the degree of the other neighbor of on , where or . Then
[TABLE]
and thus , which is also a contradiction. Thus, . If , then , and if , then . In either case, we have .
Now we have shown that the graphs in are the unique graphs in with the maximum sum–connectivity index. Note that for and ,
[TABLE]
The result follows.
Theorem 5.1
Among the graphs in with , the graph in and the graph in for are respectively the unique graphs with the maximum and the second maximum sum–connectivity indices, the graphs in and the graph in for are respectively the unique graphs with the maximum and the second maximum sum–connectivity indices, the graphs in and the graphs in for are respectively the unique graphs with the maximum and the second maximum sum–connectivity indices, where for and for .
Proof. Suppose that is a graph in with the maximum sum–connectivity index, and and are its two cycles. Let be the unique common vertex of and . By Lemma 5.1, the vertices outside and are of degree one or two, the vertices on and different from are of degree two or three, and or . Moreover, by Lemma 5.2 (i), the vertices on and different from are of degree two. If , then making use of Lemma 5.2 (iii) to , we may get a graph in with larger sum–connectivity index, a contradiction. Thus , i.e., .
Note that for , , , and ,
[TABLE]
Then the result follows from Lemmas 5.3 and 5.4.
Acknowledgement. This work was supported by the National Natural Science Foundation of China (Grant No. 10671076).
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1[1] G. Caporossi, I. Gutman, P. Hansen, Lj. Pavlović, Graphs with maximum connectivity index, Comput. Biol. Chem. 27 (2003) 85–90.
- 2[2] Z. Du, B. Zhou, N. Trinajstić, Minimum sum–connectivity indices of trees and unicyclic graphs of a given matching number, J. Math. Chem., in press.
- 3[3] Z. Du, B. Zhou, N. Trinajstić, Sum–connectivity indices of trees and unicyclic graphs of fixed maximum degree, to appear.
- 4[4] B. Furtula, A. Graovac, D. Vukičević, Atom–bond connectivity index of trees, Discrete Appl. Math. 157 (2009) 2828–2835.
- 5[5] I. Gutman, B. Furtula (Eds.), Recent Results in the Theory of Randić Index, Univ. Kragujevac, Kragujevac, 2008.
- 6[6] X. Li, I. Gutman, Mathematical Aspects of Randić–Type Molecular Structure Descriptors, Univ. Kragujevac, Kragujevac, 2006.
- 7[7] B. Lučić, N. Trinajstić, B. Zhou, Comparison between the sum–connectivity index and product–connectivity index for benzenoid hydrocarbons, Chem. Phys. Lett. 475 (2009) 146–148.
- 8[8] M. Randić, On characterization of molecular branching, J. Am. Chem. Soc. 97 (1975) 6609–6615.
