Molecular Property Prediction: A Multilevel Quantum Interactions Modeling Perspective

By Chengqiang Lu, Qi Liu, Chao Wang, Zhenya Huang, Peize Lin, Lixin He. 2019

228 citations recorded by the source corpus.

Predicting molecular properties (e.g., atomization energy) is an essential issue in quantum chemistry, which could speed up much research progress, such as drug designing and substance discovery. Traditional studies based on density functional theory (DFT) in physics are proved to be time-consuming for predicting large number of molecules. Recently, the machine learning methods, which consider much rule-based information, have also shown potentials for this issue. However, the complex inherent quantum interactions of molecules are still largely underexplored by existing solutions. In this paper, we propose a generalizable and transferable Multilevel Graph Convolutional neural Network (MGCN) for molecular property prediction. Specifically, we represent each molecule as a graph to preserve its internal structure. Moreover, the well-designed hierarchical graph neural network directly extracts features from the conformation and spatial information followed by the multilevel interactions. As a consequence, the multilevel overall representations can be utilized to make the prediction. Extensive experiments on both datasets of equilibrium and off-equilibrium molecules demonstrate the effectiveness of our model. Furthermore, the detailed results also prove that MGCN is generalizable and transferable for the prediction.

Subfields: Computational Theory and Mathematics, Materials Chemistry, Physical and Theoretical Chemistry

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Theory and Mathematics

1. Shuo Zhang
2. Yang Liu
3. Lei Xie
4. Stephan Günnemann
5. Johannes Klicpera
6. Ying Wei
7. Junzhou Huang
8. Hao Wang

In Materials Chemistry

1. Shuo Zhang
2. Yang Liu
3. Lei Xie
4. Stephan Günnemann
5. Johannes Klicpera
6. Ying Wei
7. Junzhou Huang
8. Yatao Bian

In Physical and Theoretical Chemistry

1. Stephan Günnemann
2. Johannes T. Margraf
3. Chengqiang Lu
4. S. Giri
5. Qi Liu
6. Johannes Klicpera

Citations in the corpus

• Fast and Uncertainty-Aware Directional Message Passing for Non-Equilibrium Molecules (2020, methodology): “We evaluate the final DimeNet ++ model on all QM9 targets and compare it to the state-of-the-art models SchNet [14], MGCN [9], and DeepMoleNet [8].”
• Motif-based Graph Self-Supervised Learning for Molecular Property Prediction (2021, background): “[25] builds a hierarchical GNN to capture multilevel interactions.”
• Efficient and Accurate Physics-aware Multiplex Graph Neural Networks for 3D Small Molecules and Macromolecule Complexes (2022, methodology): “From the results, we can observe that the models that incorporate only atomic pairwise distance 𝑑 as geometric information like SchNet, PhysNet, and MGCN perform worse than those models that incorporate both 𝑑 and related angles 𝜃 like HMGNN, DimeNet++, and our PaxNet.”
• Self-Supervised Graph Transformer on Large-Scale Molecular Data (2020, background): “Furthermore, [30] builds a hierarchical GNN to capture multilevel interactions.”