muxGNN: Multiplex Graph Neural Network for Heterogeneous Graphs.

Graph neural networks (GNNs) have become effective learning techniques for many downstream network mining tasks including node and graph classification, link prediction, and network reconstruction. However, most GNN methods have been developed for homogeneous networks with only a single type of node and edge. In this work we present muxGNN, a multiplex graph neural network for heterogeneous graphs. To model heterogeneity, we represent graphs as multiplex networks consisting of a set of relation layer graphs and a coupling graph that links node instantiations across multiple relations. We parameterize relation-specific representations of nodes and design a novel coupling attention mechanism that models the importance of multi-relational contexts for different types of nodes and edges in heterogeneous graphs. We further develop two complementary coupling structures: node invariant coupling suitable for node- and graph-level tasks, and node equivariant coupling suitable for link-level tasks. Extensive experiments conducted on six real-world datasets for link prediction in both transductive and inductive contexts and graph classification demonstrate the superior performance of muxGNN over state-of-the-art heterogeneous GNNs. In addition, we show that muxGNN's coupling attention discovers interpretable connections between different relations in heterogeneous networks.

Comparative effects of stressors on behavioral and neuroimmune responses of fawn-hooded (FH/Wjd) and Wistar rats: Implications for models of depression.

Patients with depression and rodent models of depression show increased cytokines and activated microglia. Fawn Hooded (FH/Wjd) rats have long been used as a model of depression based on their depressive-like behaviors, high basal corticosterone levels and altered serotonergic levels, but little is known about the neuroimmune function in this model. To test whether depressive-like behaviors relate to dysfunction of the neuroimmune system, depressive-like behaviors in the forced swim test (FST) and corticosterone (CORT) response to the swim test were compared in male Fawn Hooded versus Wistar rats, and cytokine levels in plasma and brain and plasma CORT in response to lipopolysaccharide (LPS, an endotoxin that activates the neuroimmune system) or 1 h restraint were measured. Fawn Hooded rats had more depressive-like behaviors in the FST (decreased swim time and increased immobility) and increased overall plasma CORT compared with Wistar rats. Additionally, Fawn Hooded rats exhibited blunted brain and plasma cytokine response to LPS compared with Wistar rats, an effect that might be related to the blunted plasma CORT response to LPS. No strain differences were found on these measures in response to restraint stress. These results suggest that Fawn Hooded rats have a depressive-like phenotype potentially more closely associated with serotonin dysregulation and a dysregulated HPA axis and remain a relevant model for further defining the role of these systems in depressive conditions.