Abstract: Peptides provide a powerful framework forexploring molecular function, offering routes to biomaterials, nanostructures, and bioactive assemblies. However, their immense design space makes systematic exploration a daunting task. In this talk, I will describe our efforts to merge computational chemistry with artificial intelligence to accelerate peptide discovery. I will begin with our studies on tripeptides that form soft materials, showing how machine learning can expose the hidden principles of molecular assembly. From there, I will discuss how targeted models enable the design of peptides with specific functions, including pore formation. Finally, I will present our ongoing work on using structural descriptors as design objectives and leveraging generative AI to predict peptide co-assemblies. These advances suggest new ways of approaching peptide discovery where algorithms not only guide experimental exploration but also inspire new concepts in molecular design.
Speaker Bio: Tell Tuttle is Professor of Theoretical Chemistry and Head of the Department of Pure and Applied Chemistry at the University of Strathclyde in Glasgow. His research brings together molecular simulation, machine learning, and artificial intelligence to explore peptide self-assembly and design functional nanostructures. He has published extensively on the computational discovery of minimal peptide motifs, the use of AI for targeted sequence design, and the prediction of peptide co-assemblies. As Head of Department, he also leads a large research community spanning fundamental and applied chemistry, with strong links to both industry and interdisciplinary collaboration.