Hyper-Zagreb Indices of Hypergraphs with Application in Drug Design

Let $\mathcal{H}$ be a hypergraph on the non-empty finite vertex set $V(\mathcal{H})$ with the hyperedge set $E(\mathcal{H})$, where each hyperedge $e\in E(\mathcal{H})$ is a subset of $V(\mathcal{H})$ with at least two vertices. This paper introduces the first and second Hyper-Zagreb indices for hypergraphs, extending these well-known graph indices to hypergraphs. We discuss bounds on these indices for general hypergraphs, weak bipartite hypergraphs, hypertrees, $k$-uniform hypergraphs, $k$-uniform weak bipartite hypergraphs, and $k$-uniform hypertrees, characterizing the extremal hypergraphs that achieve these bounds. Additionally, we present a novel application of these indices in drug design and bioactivity prediction, demonstrating their utility in quantitative structure-activity relationship (QSAR) modeling.