A Review of the Research Progress of Simplified Models in Protein Structure Prediction

Abstract

Protein is the main executor of life activities. Proteins form specific three-dimensional conformations through the folding of amino acid sequences. The functions of proteins are decided by their three-dimensional structure. However, the use of experimental approaches to determine protein structures has numerous limitations. Protein folding methods for prediction are constantly evolving as a result of advances in computational biology and artificial intelligence. Simplified models are frequently employed to lower computational complexity due to the complexity of the all-atomic model and the unpredictable nature of the chemical reaction. This paper discusses some core issues in protein folding prediction, reviews relevant theories such as Levinthal’s paradox, Energy Tunnel Theory, and Metadynamics, and summarizes the progress of research methods of simplified models in protein folding prediction. It was also explored how to improve the algorithm-based model. This article’s goal is to review the current status of studies on simplified models for the identification of protein structures from the perspectives of biophysical and mathematical biology.