Multi-modal property prediction method based on syntax tree constraint molecular representation

By combining context-free grammars and graph convolutional networks, a multimodal feature fusion framework is constructed, which overcomes the limitations of existing molecular property prediction models in parsing chemical syntax rules, and achieves more efficient and accurate molecular property prediction, supporting drug design.

CN121415917APending Publication Date: 2026-01-27NANTONG UNIV
0 Cites 0 Cited by

Patent Information

Application Number
CN202511615543.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-11-06
Publication Date
2026-01-27

AI Technical Summary

Technical Problem

In the existing technology, molecular property prediction models based on sequence features or graph structure features have limitations in resolving nested molecular substructures and chemical grammar rules, resulting in prediction results that lack chemical rationality and accuracy, especially in drug design.

Method used

We combine context-free syntax tree representation with molecular property prediction tasks. We capture the sequence dependencies of SMILES and the topological structure of molecular graphs through Transformer and GCN respectively, and use channel splicing strategy to achieve cross-dimensional feature interaction, thus constructing a comprehensive representation that includes chemical semantics, spatial structure and syntactic information.

Benefits of technology

It improves the accuracy and reliability of molecular property prediction, provides interpretable chemical structure clues, and enhances the precision of drug design and the ability to capture chemical semantics.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention discloses a multi-modal property prediction method based on syntax tree constraint molecular representation. The method comprises the following steps: combining syntax tree representation based on context-independent grammar with a molecular property prediction task; setting a multi-modal feature fusion framework, and establishing strong relevance between features and prediction results; and the fusion features are finally input into the predictor network module to complete the prediction of the properties of the target molecules. According to the method, syntax tree representation based on context-independent grammar is combined with a molecular property prediction task, a hierarchical structure and a chemical grammar rule of coding molecules are displayed, it is ensured that prediction molecules meet SMILES grammar specifications, and the problem that grammar constraints of a traditional sequence model are missing is fundamentally solved.
Need to check novelty before this filing date? Find Prior Art