• proteins
• systems biology
• machine learning
• artificial intelligence
• dynamical systems
• supervised learning
• graph neural networks

This thesis focuses on the application of machine learning techniques to explore the relationship between biochemical pathways and protein-protein interaction (PPi) networks in biological systems. PPi networks, with limited information compared to pathways, present challenges in inferring relevant properties on the emerging behavior of a complex system. The work demonstrates that the dynamical information coming from the pathway level can still be predicted on PPi networks using graph neural networks.