Our colleague Adrian Rošinec recently attended the BioExcel conference, showcasing a poster on the challenges and opportunities in molecular dynamics (MD) data management. His presentation highlighted the limitations of metadata in identifying and comparing complex MD trajectories, emphasizing the need for advanced similarity search methods.
The poster explored how embedding techniques can transform MD trajectories into compact, searchable vectors, enabling comparisons that reveal functional similarities and dynamic behaviours beyond metadata. Adrian's work, inspired by tools like AlphaFind [1,2], aims to advance the MD community’s ability to effectively analyze and interpret large-scale simulation data.
PDBCharges, a tool providing partial atomic charges for protein structures from the Protein Data Bank, was published in the journal Nucleic Acids Research.
