Clara (Chen) Shao
PhD Candidate in Computer Science, Karlsruhe Institute of Technology (KIT)
cshao676 [AT] gmail.com
Graph Neural Networks for Materials & Chemistry
Research in the AiMat Group, KIT
This line of work was carried out in the AiMat group (Artificial Intelligence for Materials Science) led by Prof. Pascal Friederich at KIT’s Institute of Theoretical Informatics, applying graph neural networks to problems in chemistry and materials science.
Master’s Thesis — Energy Difference Prediction in Chemistry Using Graph Neural Networks. Supervised by Prof. Pascal Friederich (AiMat, ITI), my thesis developed graph neural network models to predict energy differences between molecular configurations, exploring how to learn joint representations of disjoint molecular graphs for chemical property prediction.
Related publications from this research:
- Review of Graph Neural Networks for Materials — a comprehensive survey of GNN methods for materials science, published in Communications Materials (3, 93, 2022).
- Graph Neural Networks to Learn Joint Representations of Disjoint Molecular Graphs (with Chen Zhou and Pascal Friederich, arXiv:2210.09517, 2022).
This early work in graph machine learning for the natural sciences laid the foundation for my ongoing research on graphs, and connects directly to my current work on foundation models for protein design.