Graph neural networks have achieved impressive results in predicting molecular properties, but they do not directly account for local and hidden structures in the graph such as functional groups and molecular geometry. At each propagation step, graph neural networks aggregate only over first order neighbours and can only learn about important information contained in subsequent neighbours as well as the relationships between those higher order connections—over many propagation steps. In this work, we generalize graph neural nets to pass messages and aggregate across higher order paths. This allows for information to propagate over various levels and substructures of the graph. We demonstrate our model on a few tasks in molecular property prediction.

图神经网络在预测分子特性方面取得了令人瞩目的成果，但它们并没有直接解释图中的局部和隐藏结构，例如官能团和分子几何结构。在每个传播步骤中，图神经网络仅聚合一阶邻居，并且只能在许多传播步骤中了解包含在后续邻居中的重要信息以及这些高阶连接之间的关系。在这项工作中，我们将图神经网络泛化为跨高阶路径传递消息和聚合。这允许信息在图的各个级别和子结构上传播。我们在分子特性预测中的一些任务上展示了我们的模型。