Periodic Reporting for period 1 - SupraModel (Peptide-based Supramolecular Co-assembly Design: Multiscale Machine Learning Modeling Approach)
Período documentado: 2023-04-01 hasta 2025-09-30
Peptides, short chains of amino acids, can spontaneously assemble into multi-functional materials used in emerging technologies ranging from drug delivery to soft semiconductor devices. Designing these materials is challenging because scientists must identify the best amino acid sequence among enormous possibilities, much larger than one can experimentally test. Thus, there is an urgent need to understand the underlying principles that will allow for a shift from a human intuition-based to a rational design-based approach. The EU funded SupraModel project aims to develop an efficient and rapid framework for predicting peptide co-assembly structures while providing unprecedented molecular insight into the co-assembly process.
Novel strategies were developed to enhance the accuracy and efficiency of Graph Neural Network-based molecular models. These methods were implemented in new, well-documented software tools, reducing the entry barrier to the state-of-the-art simulation techniques for a general molecular modeling community.
Accurate and efficient molecular models of organic molecules are essential to advancing many scientific fields, ranging from medicine to environmental sciences.