Graph classification method based on attention mechanism and compound toxicity prediction method

Abstract

The invention discloses a graph classification method based on an attention mechanism. The method comprises the steps of obtaining an original graph network and corresponding graph network data information; initializing feature embedding of nodes, constructing a high-order graph network, and initializing node embedding in the high-order graph network; capturing and absorbing neighbor node information and interaction information between neighbor nodes by nodes in the graph network of each order by adopting a substructure interaction attention network, and training and obtaining new feature embedding of each node in the graph network; adopting a node attention network to embed and fuse node features in the graph network into graph network embedding, and after cascading each order of graph network embedding, performing dimension reduction classification through a multi-layer perceptron to obtain a final graph network classification result. The invention also discloses a compound toxicity prediction method comprising the attention mechanism-based graph classification method. According to the method, the accuracy of graph network classification and the accuracy of compound toxicity prediction are effectively improved; and the method is relatively high in efficiency, relatively good in accuracy and easy to implement.

Description

technical field [0001] The invention belongs to the field of graph data mining, and in particular relates to a graph classification method based on an attention mechanism and a compound toxicity prediction method. Background technique [0002] As an important branch in the field of graph mining, graph classification has been applied to many practical fields. Judging whether the DNA protein sequence is mutated, predicting whether an unknown compound is toxic, and classifying the social group based on the internal structural attributes of the social group can all be abstracted into graph classification problems. For example, in the field of chemistry, chemists use graphs to model the structure of compounds, use the vertices in the graph to represent the molecules in the compound, and use the edges in the graph to represent the chemical bonds between molecules. By mining the compound atlas, Intrinsic relationships between compound structures can be discovered, helping to find ...

AI Technical Summary

Problems solved by technology

But these models have shortcomings: on the one hand, GAT and its variants only focus on a single node (first-order), which limits their accuracy in graph classification applications; on the other hand, these high-order models only use adjacency matrices or The variant linearly absorbs the information of neighbors, and has difficulties in dealing with complex nonlinear data (such as mutual information between neighbors); and these nonlinear data cannot be directly absorbed by existing convolution operations, but the information they contain will Affects the representation of graph information, thus causing the model to be flawed in accuracy

Images