When artificial intelligence meets protein research

Abstract

The 2024 Nobel Prizes in Chemistry and Physics mark a watershed moment in the convergence of artificial intelligence (AI) and molecular biology. This article explores how AI, particularly deep learning and neural networks, has revolutionized protein science through breakthroughs in structure prediction and computational design. It highlights the contributions of 2024 Nobel laureates John Hopfield, Geoffrey Hinton, David Baker, Demis Hassabis, and John Jumper, whose foundational work laid the groundwork for AI tools such as AlphaFold. These tools are transforming our understanding of protein folding, and the dynamics of non-globular proteins, including intrinsically disordered proteins. While AI-driven methods have made predicting protein structures faster and more accessible, they also underscore ongoing scientific challenges, including the dynamics of protein folding and amyloid aggregation. European initiatives, such as the COST Actions NGP-net (BM1405) and ML4NGP (CA21160), are spearheading efforts to bridge these gaps by integrating AI and experimental data in the study of non-globular proteins. Together, these developments signal a transformative shift in biology, paving the way for novel discoveries in medicine, biotechnology, and materials science.

Keywords: artificial intelligence; bioinformatics; deep learning; machine learning; non-globular proteins; protein folding; structural biology.