The study of the dynamic behavior of proteins is crucial for understanding many biological mechanisms, such as enzymatic catalysis and allosteric regulation. While experimental methods provide a solid foundation for these dynamic studies, Molecular Dynamics (MD) simulations allow us to analyze these motions with atomic-level details. Thanks to recent technical advances, we can now simulate larger systems of up to a million atoms over longer time scales, with microseconds becoming the new standard. One of the biggest challenges is analyzing these simulations, which constitute a complex system. Network theory is particularly appealing to simplify the complex information contained in 3D protein structures. Furthermore, network theory methods such as centrality measures, modularity calculations, and connected component analysis offer endless possibilities for characterizing the dynamic behavior of each amino acid. 

This presentation will explore how network theory can be applied to structural big data to address and simplify various problems in structural bioinformatics. First, we will begin with global structural comparisons that allowed us to study allosteric mechanisms and conformational rearrangements caused by mutations. Then, we will move on to more local comparisons that enabled us to identify evolutionary convergences in the structure of binding sites. Finally, we will demonstrate how these methods can be applied to the field of enzyme design.