Machine Learning (ML) has become an essential tool in modern structural biology, offering unprecedented opportunities to analyse, model, and predict the behaviour of non-globular proteins (NGPs), including intrinsically disordered proteins and regions, repeat proteins, and other flexible systems. The integration of ML approaches with biophysical and computational data is reshaping our understanding of protein disorder, dynamics, and function.

The aim of this Training School is to provide participants with both theoretical foundations and practical experience in applying ML methods to NGP research. Through a balanced combination of lectures and hands-on sessions, participants will explore how different ML techniques—ranging from classical algorithms to deep learning architectures—can be employed to analyse NGP sequences, predict disorder-related features, and interpret model outputs. Emphasis will be placed on understanding the underlying principles, proper evaluation of results, and identification of potential pitfalls and biases in ML applications.

The Training School is designed for PhD students, postdoctoral researchers, and young investigators with a background in life sciences, bioinformatics, or related disciplines. Participants will gain the conceptual and technical skills needed to interpret ML results critically and to apply them effectively in their own research. By fostering active participation and discussion, this event aims to strengthen the growing community of researchers working at the intersection of machine learning and protein science, advancing our collective understanding of non-globular proteins.