A review of AlphaFold, a method for protein structure prediction

Abstract

Introduction: Proteins are regarded as one of the most important biological macromolecules. Identifying their structure will help researchers understand protein’s function and practical applications. In the past, experimental and classical methods such as X-ray crystallography or NMR were used, with advantages and disadvantages. Today, new methods based on artificial intelligence have made a big change in predicting the structure of proteins, an example of which is the method called AlphaFold. Methodology: Articles related to the present topic were reviewed from databases such as Google Scholar and Pubmed from 2019 to 2024. Results: AlphaFold, a method for predicting the structure of proteins in three versions has been introduced to the world. This method, based on deep learning based on convolutional neural networks, processes input data that are often collected from PDB and accurately and quickly predicts the structure of proteins. This method can be widely used in medical, biological, education, industrial and production cases. Discussion and conclusion: Different methods are used in protein structure prediction. New methods that are based on artificial intelligence have attracted the attention of researchers because they have high accuracy and speed. However in order to gain confidence, they need to be measured.