Protein-protein interactions (PPIs) play key roles in a broad range of biological processes. Numerous strategies have been proposed for predicting PPIs, and among them, graph-based methods have demonstrated promising outcomes owing to the inherent graph structure of PPI networks. This paper reviews various graph-based methodologies, and discusses their applications in PPI prediction. We classify these approaches into two primary groups based on their model structures. The first category employs Graph Neural Networks (GNN) or Graph Convolutional Networks (GCN), while the second category utilizes Graph Attention Networks (GAT), Graph Auto-Encoders and Graph-BERT. We highlight the distinctive methodologies of each approach in managing the graph-structured data inherent in PPI networks and anticipate future research directions in this domain.

We investigate the effects of charged-current (CC) nonstandard neutrino interactions (NSIs) at the source and at the detector in the simulated data for the planned Deep Underground Neutrino Experiment (DUNE), while neglecting the neutral-current NSIs at the propagation due to the fact that several solutions have been proposed to resolve the degeneracies posed by neutral-current NSIs while no solution exists for the degeneracies due to the CC NSIs. We study the effects of CC NSIs on the simultaneous measurements of $\theta_{23}$ and $\delta_{CP}$ in DUNE. The analysis reveals that 3$\sigma$ C.L. measurement of the correct octant of $\theta_{23}$ in the standard mixing scenario is spoiled if the CC NSIs are taken into account. Likewise, the CC NSIs can deteriorate the uncertainty of the $\delta_{CP}$ measurement by a factor of two relative to that in the standard oscillation scenario. We also show that the source and the detector CC NSIs can induce a significant amount of fake CP-violation and the CP-conserving case can be excluded by more than 80\% C.L. in the presence of fake CP-violation. We further find the potential of DUNE to constrain the relevant CC NSI parameters from the single parameter fits for both neutrino and antineutrino appearance and disappearance channels at both the near and far detectors. The results show that there could be improvement in the current bounds by at least one order of magnitude at the near and far detector of DUNE except a few parameters which remain weaker at the far detector.