Advances of computational protein design: Principles, strategies and applications in nutrition and health

Abstract

Computational methods and AI technology have had a profound impact on protein design, significantly enhancing the ability to predict protein structures and create proteins with custom-tailored functions. With the help of computational methods, traditional protein design strategies such as directed evolution, fusion protein, and key subunit interface redesign show unprecedented progress in the design of various protein biomaterials such as nanocages, nanocarriers, antibodies, biocatalytic enzymes and inhibitory peptides. Strategies include physics-mediated design, which leverages the physical principles underlying protein structure and dynamics, and AI-mediated design, which employs machine learning techniques to generate and optimize protein configurations. Together, these approaches represent the cutting-edge methodologies in the rational design of novel proteins with desired functions. By using these approaches, novel protein molecules, assemblies, antibodies and responsive nanofibrils were constructed, which can be further applied in the field of nutrition and health.