Chemical graph theory studies the modeling of chemical structures by graphs. For this purpose, many different graph invariants have been defined. One of the most well-known invariants is the Wiener index. In certain cases, a graph can be isometrically embedded into the $\ell_1$-space, which provides useful properties for the computation of various indices. Over the years, the use of $\ell_1$-embeddings for this purpose has become known as the cut method. In this dissertation, we address related problems in the context of hypergraphs. Recently, an increasing number of studies have focused on the computation of the Wiener index for hypergraphs. In this work, we extend the cut method to hypergraphs and demonstrate its application, primarily for the calculation of the Wiener index. To this end, we introduce the notions of cube hypergraphs and partial cube hypercubes, and present generalizations of classical theorems from graph theory to the hypergraph setting. These generalizations include a characterization of partial cube hypercubes and the canonical metric embedding of hypergraphs.
We further study various operations on hypergraphs that generate new structures by adding vertices or hyperedges, and analyze the $\ell_1$-embeddability of the resulting hypergraphs. Moreover, we investigate the Wiener index of these new hypergraphs and, whenever possible, establish relations between the indices of the original and the resulting hypergraphs. We also show that such operations allow for the construction of hypergraphs naturally motivated by chemical graph theory.
Special attention is devoted to several concrete families of hypergraphs, such as sunflowers, tight hyperpaths, and tight hypercycles. In addition, we examine chemically motivated families of hypergraphs, including phenylene hypergraphs, Clar hypergraphs, and closed neighborhood hypergraphs.
Finally, we introduce definitions of other indices for hypergraphs and demonstrate that the developed methods can be adapted to these cases as well. We find that, apart from the Wiener index, other indices have not yet been formally defined or studied in the context of hypergraphs. To this end, we define the Szeged and $PI$ indices for hypergraphs and formulate a basic version of the cut method for their computation.
|