![MathJax Logo](/templates/jsp/_style2/_tandf/pb2/images/math-jax.gif)
Abstract
The revised edge Szeged index of a graph G is defined as
where
and
are, respectively, the number of edges of G lying closer to vertex u than to vertex v and the number of edges of G lying closer to vertex v than to vertex u, and
is the number of edges equidistant to u and v. In the paper, we show the sharp lower bound of revised edge Szeged index regarding bicyclic graphs. Moreover, all extremal graphs are characterized.
1. Introduction
All graphs considered in this paper are finite, undirected and simple. We refer the readers to [Citation4] for terminology and notations. Let Cn denote the cycle on n edges and Sm denote the star on m edges. For denotes the distance between u and v in G. The Wiener index of a connected graph G is defined as follows:
This topological index has been extensively studied in the mathematical literature, see [Citation10, Citation14, Citation15, Citation30].
Let e = uv be an edge of G, the sets and
are the set of vertices equidistant from u and v, the set of vertices whose distance to vertex u is smaller than the distance to vertex v and the set of vertices closer to v than u, respectively. The number of vertice of
and
are denoted by
and
respectively. Evidently,
For a tree T, its Wiener index can be defined alternatively as:
Motivated the above formula, Gutman [Citation12] introduced a graph invariant, named as the Szeged index as an extension of the Wiener index and defined by
In [Citation25], Randić noticed that the Szeged index neglects the vertices with equal distance to the two endpoints of some edge. Hence he put forward a modified version of the Szeged index as revised Szeged index, which is defined as
So far, many results explored the relationship (difference or quotient) between Szeged index (resp. revised Szeged index) and Wiener index; see, [Citation3, Citation6, Citation8, Citation9, Citation19, Citation22, Citation23, Citation33, Citation34]. There are a lots results with the bounds and applications of Szeged indices, refer to [Citation1, Citation2, Citation7, Citation17, Citation18, Citation20, Citation24, Citation26, Citation29, Citation31, Citation32].
Given an edge the distance between the edge e and the vertex w, denoted by d(e, w), is defined as
Based on the definition, the three subsets of E(G) are proposed as
Put
and
Then, the edge Szeged index [Citation13], and revised edge Szeged index [Citation11] of G are defined as follows:
Results on edge Szeged index can be found in [Citation5, Citation16, Citation27, Citation28]. In [Citation11], they determined the n-vertex unicyclic graphs with the largest and the smallest revised edge Szeged indices. In [Citation21], they identified those graphs whose revised edge Szeged index is maximal among bicyclic graphs. In the paper, we show the sharp lower bound of revised edge Szeged index regarding bicyclic graphs and characterize extremal graphs.
Theorem 1.1.
Let G be a connected bicyclic graph of size . Then
with equality if and only if
, where B is depicted in .
Figure 1. These graphs are used in Theorem 1.1 and Theorem 3.1.
![Figure 1. These graphs are used in Theorem 1.1 and Theorem 3.1.](/cms/asset/fa0b7de5-0785-4510-b396-4fab998568fc/uakc_a_2120440_f0001_b.jpg)
2. Some Lemmas
Note that it is deduced that
Set where e = uv, we get a normal version of revised edge Szeged index
(1)
(1)
In the section, some elementary results are listed, which will be useful in the sequel.
Lemma 2.1.
Let . Then
and equality holds only if e is a pendant edge.
We now mark where
is the graph obtained from the G1 and G2 by identifying two vertices. Set w be the common vertex of G1 and G2. Obviously, w is a cut vertex of G. For every
w belongs to one of the three sets
Since every path connecting u(v) and each vertex in
is via w, all edges of G2 must contained in one of the three sets
(the same with w). Therefore, the contribution of G2 to the item
completely relies on the size of G2, that is, changing the structure of G2 cannot alter the value
Since the contribution of the pendant edge to
is the smallest, we have the following lemma.
Lemma 2.2.
Let be a connected graph of size m. Then
where the common vertex of
is the center vertex of Sm.
Let where the common vertex w of
is the center vertex of
we called w is the center vertex of
Lemma 2.3.
[Citation11] For , let G be an
edges unicyclic graph. Then
with equality if and only if
for
or
for m = 15, and
for
Lemma 2.4.
Let H1 be a connected graph of size m1 and H2 be an unicyclic graphs of sizze Then
for
for
, and
for
, where the common vertex of
is the center vertex of
Proof.
If Then, by Lemma 2.3, we get
Samely, if
then
and if
then
□
3. Proof of Theorem 1.1
A Θ-graph is a graph that connecting two fix vertices by three internally disjoint paths. If the corresponding three paths have the lengths a, b, c with we denote it by
Let
be the set of m-edge bicyclic connected graphs with exactly two cycles. Let
be the set of m-edge bicyclic connected graphs with three cycles. That is if
then it has a Θ graph as its subgraph, which is called the brace of G.
From Lemma 2.4, we deduce the following conclusion.
Theorem 3.1.
Let . If
, then
with equality only if
; If m = 15, then
with equality only if
, where
is depicted in ; If
, then
with equality only if
Proof.
Since there are two unicyclic graphs H1 and H2 such that
Assume that
In view of Lemma 2.4, we get
for
Clearly, and
Therefore, the proof is complete. □
Lemma 3.2.
Let with brace
. If
, then
and equality holds only if
, where B2 is depicted in .
Proof.
Suppose that x1 and x2 are the vertices with degree 3 and are the vertices with degree 2 in
of G. In view of Lemma 2.2, we have that each edge in the pendant tree of each xi is a pendant edge. Let li be the number of pendant edges of xi. Assume that
by symmetry. G1 is the graph from G by transferring l2 pendant edges from x2 to x1. It is checked that
It follows that and equality holds only if
G2 is the graph from G1 by moving l5 pendant edges from x5 to x3. It is deduced that
By means of EquationEq. (1)(1)
(1) ,
with equality only if
G3 denotes the graph from G2 by shifting l4 pendant edges from x4 to x3. So
by symmetry, and equality holds only if
G4 is the graph from G3 by transferring l1 pendant edges from x1 to x3. Clearly, Moreover,
Using EquationEq. (1)(1)
(1) , we get
and equality holds only if
Moreover,
Hence the proof is finished. □
Lemma 3.3.
Let with brace
. If
, then
with equality only if
; If m = 9, then
with equality only if
, where B3, B4 is depicted in ; If
, then
with equality only if
Proof.
Let x1 and x2 be two vertices with degree 3 and x3, x4 are three vertices with degree 2 in of G. In view of Lemma 2.1, we have that each edge in the pendant tree of each xi is a pendant edge. Let li be the number of pendant edges of xi. Suppose
from symmetry. G1 is the graph from G by moving l2 pendant edges from x2 to x1. Moreover, we arrive at
Using EquationEq. (1)(1)
(1) , we show
with equality only if
Let G2 deonte the graph from G1 by shifting l4 pendant edges from x4 to x3. Then,
Combining with EquationEq. (1)(1)
(1) ,
with equality only if
G3 is the new graph obtained from G2 by moving l1 pendant edges from x1 to x3. Hence, Furthermore,
Note that with
So
by EquationEq. (1)
(1)
(1) , and equality holds only if
Clearly,
with
In addition,
Our proof is finished. □
Lemma 3.4.
Let with brace
. If
, then
and equality holds only if
, where B5 is depicted in .
Proof.
Assume that x1 and x2 are two vertices with degree 3 and are three vertices with degree 2 of
in G. In view of Lemma 2.1, we have that each edge in the pendant tree of each xi is a pendant edge. Let li be the number of pendant edges of xi. Assume that
from symmetry. G1 denotes the graph from G by transferring l2 pendant edges from x2 and l5 pendant edges of x5 to x1 and x4, respectively. We deduce that
Combining with EquationEq. (1)(1)
(1) ,
and equality holds only if
G2 denotes the graph obtained from G1 by moving l4 pendant edges from x4 to x1. Then, we have
Using EquationEq.(1)(1)
(1) ,
with equality only if
G3 is the graph obtained from G2 by shifting l3 pendant edges from x3 to x1. Clearly, Moreover,
Identifying with EquationEq.(1)
(1)
(1) ,
and equality holds with only if
In addition, We complete the proof. □
Theorem 3.5.
Let with brace
. If
, then
with equality only if
; If
, then
with equality only if
; If m = 9, then
with equality only if
; If
, then
with equality only if
Proof.
Let be the three paths of
in G. Mark x and y the two vertices with degree 3 of
In order to complete the proof, we take the following three cases.
Case 1.
We choose all edges with at least one end of x and y in
e = xz denotes one of the six edges. Note that
for
and
otherwise, then we have
This fact also holds for other five edges. So
for
Consisting with EquationEq.(1)
(1)
(1) , the Case is true.
Case 2.
Subcase 2.1.
The Subcase follows from Lemma 3.2.
Subcase 2.2.
Take four edges of the paths and
Set one edge of them as e = xz, clearly,
Choose the two edges of the path
satisfying incident with x or y. Put down one of them as e = xw, obviously,
Hence
By means of EquationEq. (1)(1)
(1) , the Subcase is true.
Subcase 2.3.
Using the similar discussion of Subcase 2.2, the Subcase is also true.
Case 3.
Subcase 3.1.
From Lemma 3.3, we deduce that if then
If
then
If m = 9, then
If m = 6, 7, 8, then
If m = 5, then
Consequently, the Subcase is true.
Subcase 3.2.
The Subcase follows from Lemma 3.4.
Subcase 3.3.
Take the five edges that are with at least one end of x and y in Set one edge e = xy, then
set e = xz from one of the other four edges, obviously,
In addition, take the two edges from the middle of
e = uv denotes one of the two edges. we get
Hence
for
The Subcase is true by EquationEq.(1)
(1)
(1) .
Subcase 3.4.
Choose the five edges that are with at least one end of x and y of If e = xy, then
e = xz denotes one of other four edges, then
Take two edges from the middle of
Let e = uv be one of them. Then
Hence
Together with EquationEq. (1)
(1)
(1) , the Subcase is true.
Therefore, we finish the proof. □
From Theorem 3.1 and Theorem 3.5, we not only show the Theorem 1.1, but also obtain a strengthening result.
Theorem 3.6.
Let G be a bicyclic graph G with size . Then
Acknowledgements
M. Liu is supported by National Natural Science Foundation of China (No. 11961040), Innovation ability improvement project of colleges and universities in Gansu Province (No. 2019A-037), Tianyou Youth Talent Lift Program of Lanzhou Jiaotong University. S. Ji is supported Shandong Provincial Natural Science Foundation of China (No. ZR2019MA012).
Disclosure statement
No potential conflict of interest was reported by the authors.
Data availability
The data used to support the findings of this study are available from the corresponding author upon request.
Additional information
Funding
References
- Aouchiche, M, Hansen, P. (2010). On a conjecture about the Szeged index. Eur. J. Combin. 31(7): 1662–1666.
- Azari, M. (2017). Some variants of the Szeged index under rooted product of graphs. MMN 17(2): 761–775.
- Bonamy, M., Knor, M., Lužar, B., Pinlou, A, Škrekovski, R. (2017). On the difference between the Szeged and Wiener index. Appl. Math. Comput. 312: 202–213.
- Bondy, J. A, Murty, U. S. R. (2008). Graph Theory. Berlin: Springer-Verlag.
- Cai, X, Zhou, B. (2010). Edge Szeged index of Unicyclic graphs. MATCH Commun. Math. Comput. Chem. 63: 133–144.
- Chen, L., Li, X, Liu, M. (2014). The (revised) Szeged index and the Wiener index of a nonbipartite graph. Euro. J. Comb. 36: 237–246.
- Chen, L., Li, X, Liu, M. (2014). Tricyclic graphs with maximal revised Szeged index. Discr. Appl. Math. 177: 71–79.
- Chen, L., Li, X., Liu, M, Gutman, I. (2012). On a relation between the Szeged index and the Wiener index for bipartite graphs. Trans. Comb. 1(4): 43–49.
- Črepnjak, M, Tratnik, N. (2017). The Szeged index and the Wiener index of partial cubes with applications to chemical graphs. Appl. Math. Comput. 309: 324–333.
- Dobrynin, A. A., Entringer, R, Gutman, I. (2001). Wiener index of trees: theory and applications. Acta Appl. Math. 66(3): 211–249.
- Dong, H., Zhou, B, Trinajstić, C. (2011). A novel version of the edge-Szeged index. Croat. Chem. Acta 84: 543–545.
- Gutman, I. (1994). A formula for the Wiener number of trees and its extension to graphs containing cycles. Graph Theory Notes N. Y. 27: 9–15.
- Gutman, I, Ashrafi, A. R. (2008). The edge version of the Szeged index. Croat. Chem. Acta 81: 263–266.
- Gutman, I., Klavžar, S, Mohar, B. (1997). Fifty years of the Wiener index. MATCH Commun. Math. Comput. Chem 35: 1–259.
- Gutman, I., Yeh, Y. N., Lee, S. L, Luo, Y. L. (1993). Some recent results in the theory of the Wiener number. Indian J. Chem. 32A: 651–661.
- Khalifeh, M. H., Yousefi-Azari, H., Ashrafi, A. R, Gutman, I. (2008). The edge Szeged index of product graphs. Croat. Chem. Acta 81: 277–281.
- Khadikar, P., Kale, P., Deshpande, N., Karmarkar, S, Agrawal, V. (2000). Szeged indices of hexagonal chains. MATCH Commun. Math. Comput. Chem. 43: 7–15.
- Khadikar, P., Karmarkar, S., Agrawal, V. K., Singh, J., Shrivastava, A., Lukovits, I, Diudea, M. V. (2005). Szeged index-Applications for drug modeling. LDDD 2(8): 606–624.
- Li, S, Zhang, H. (2017). Proofs of three conjectures on the quotients of the (revised) Szeged index and the Wiener index and beyond. Discr. Math. 340(3): 311–324.
- Li, X, Liu, M. (2013). Bicyclic graphs with maximal revised Szeged index. Discr. Appl. Math. 161(16–17): 2527–2531.
- Liu, M, Chen, L. (2016). Bicyclic graphs with maximal edge revised Szeged index. Discr. Appl. Math. 215: 225–230.
- Nadjafi-Arani, M. J., Khodashenas, H, Ashrafi, A. R. (2011). On the differences between Szeged and Wiener indices of graphs. Discr. Math. 311(20): 2233–2237.
- Nadjafi-Arani, M. J., Khodashenas, H, Ashrafi, A. R. (2012). Graphs whose Szeged and Wiener numbers differ by 4 and 5. Math. Comput. Model. 55(3–4): 1644–1648.
- Pisanski, T, Randić, M. (2010). Use of the Szeged index and the revised Szeged index for measuring network bipartivity. Discr. Appl. Math. 158(17): 1936–1944.
- Randić, M. (2002). On generalization of Wiener index for cyclic structures. Acta Chim. Slov. 49: 483–496.
- Simić, S., Gutman, I, Baltić, V. (2000). Some graphs with extremal Szeged index. Math. Slovaca 50: 1–15.
- Tratnik, N. (2017). The edge-Szeged index and the PI index of benzenoid systems in linear time. MATCH Commun. Math. Comput. Chem. 77: 393–406.
- Vukičević, D. (2009). Note on the graphs with the greatest edge-Szeged index. MATCH Commun. Math. Comput. Chem. 61: 673–681.
- Wang, S. (2017). On extremal cacti with respect to the Szeged index. Appl. Math. Comput. 309: 85–92.
- Wiener, H. (1947). Structural determination of paraffin boiling points. J. Am. Chem. Soc. 69(1): 17–20.
- Xing, R, Zhou, B. (2011). On the revised Szeged index. Discr. Appl. Math. 159(1): 69–78.
- Žerovnik, J. (1999). Szeged index of symmetric graphs. J. Chem. Inf. Comput. Sci. 39(1): 77–80.
- Zhang, H., Chen, J, Li, S. (2017). On the quotients between the (revised) Szeged index and Wiener index of graphs. Discr. Math. Theor. Comput. Sci. 19(1): 12–25.
- Zhang, H., Li, S, Zhao, L. (2016). On the further relation between the (revised) Szeged index and Wiener index. Discr. Appl. Math. 206: 152–164.