Knowledge-enhanced Relation Graph and Task Sampling for Few-shot Molecular Property Prediction

Zeyu Wang,Tianyi Jiang,Yao Lu,Xiaoze Bao,Shanqing Yu,Bin Wei,Qi Xuan

2024-05-24

Abstract:Recently, few-shot molecular property prediction (FSMPP) has garnered increasing attention. Despite impressive breakthroughs achieved by existing methods, they often overlook the inherent many-to-many relationships between molecules and properties, which limits their performance. For instance, similar substructures of molecules can inspire the exploration of new compounds. Additionally, the relationships between properties can be quantified, with high-related properties providing more information in exploring the target property than those low-related. To this end, this paper proposes a novel meta-learning FSMPP framework (KRGTS), which comprises the Knowledge-enhanced Relation Graph module and the Task Sampling module. The knowledge-enhanced relation graph module constructs the molecule-property multi-relation graph (MPMRG) to capture the many-to-many relationships between molecules and properties. The task sampling module includes a meta-training task sampler and an auxiliary task sampler, responsible for scheduling the meta-training process and sampling high-related auxiliary tasks, respectively, thereby achieving efficient meta-knowledge learning and reducing noise introduction. Empirically, extensive experiments on five datasets demonstrate the superiority of KRGTS over a variety of state-of-the-art methods. The code is available in this https URL.

Quantitative Methods,Artificial Intelligence,Machine Learning

What problem does this paper attempt to address?

### What problem does this paper attempt to solve?

This paper aims to address the Few-shot Learning (FSL) problem in Molecular Property Prediction (MPP). Specifically:

1. **Challenges in Few-shot Molecular Property Prediction (FSMPP)**:
- Existing methods often overlook the many-to-many relationships between molecules and properties, which limits their performance.
- Similar molecular substructures can inspire the exploration of new compounds.
- The correlation between different properties can be quantified, and highly correlated properties provide more information when exploring target properties compared to lowly correlated properties.

2. **Proposed Method**:
- The paper proposes a new meta-learning framework called KRGTS (Knowledge-enhanced Relation Graph and Task Sampling), which includes a knowledge-enhanced relation graph module and a task sampling module.
- The knowledge-enhanced relation graph module constructs a Molecular-Property Multi-Relation Graph (MPMRG) to capture the many-to-many relationships between molecules and properties.
- The task sampling module includes a meta-training task sampler and an auxiliary task sampler, responsible for scheduling the meta-training process and sampling highly correlated auxiliary tasks.

3. **Main Contributions**:
- Constructed a knowledge-enhanced Molecular-Property Relation Graph (MPMRG) based on molecular substructure similarity and property information, effectively capturing the many-to-many relationships between molecules and properties.
- Proposed a method to sample highly correlated auxiliary tasks to explore target molecular properties.
- Extensive experiments demonstrate that KRGTS outperforms various state-of-the-art techniques across multiple datasets.

Through these methods, the paper aims to improve the performance of few-shot molecular property prediction, especially in scenarios with limited data.