“…This unstructured nature enables GNNs to naturally handle a wide range of graph analytics problems (i.e., node classification, link prediction, data visualization, graph clustering community detection, anomaly detection) and have been applied effectively across a diverse range of domains, e.g., protein structure prediction (Jumper et al, 2021), untangling the mathematics of knots (Davies et al, 2021), brain networks (Rosenthal et al, 2018;Xu et al, 2020b,0) in brain imaging, molecular networks (Liu et al, 2019) in drug discovery, protein-protein interaction networks (Kashyap et al, 2018) in genetics, social networks (Wang et al, 2019b) in social media, bank-asset networks (Zhou and Li, 2019) in finance, and publication networks (West et al, 2016) in scientific collaborations.…”