Protein structure prediction method based on residue distance enhanced search

Abstract

A protein structure prediction method based on residue distance enhanced search comprises the following steps: under a basic framework of a memetic algorithm, firstly, performing population initialization; secondly, performing crossover variation on a loop region of each conformation in the population to increase population diversity; then, carrying out local enhancement on the conformation by utilizing a Minmover protocol of Rosetta, and carrying out survival selection on the population according to energy and residue-residue distance distribution potential energy, so as to updating the population; and finally, outputting the final generation of population. According to the protein structure prediction method based on residue distance enhanced search, the problem of inaccurate energy function can be relieved, the conformation diversity can be increased, and the prediction precision is improved while the sampling efficiency is improved. According to the method, the defect that conformations are evaluated only through a single energy function is overcome, the population diversity is increased, and therefore the overall prediction precision is improved.

Description

technical field [0001] The invention relates to the fields of bioinformatics and computer applications, in particular to a protein structure prediction method based on residue distance enhanced search. Background technique [0002] Protein is the basic element of living cells and has a series of biological functions, including gene regulation, metabolic regulation, movement and support for the body, oxygen transport in the blood, and iron storage. The three-dimensional structure of any protein is one of the major determinants of its unique functional properties. Determining protein structure is therefore a fundamental step in understanding the function of important building blocks of life. [0003] At present, the three-dimensional structure of proteins stored in the protein structure database is mainly obtained by X-ray crystallography, nuclear magnetic resonance imaging and cryo-electron microscopy. However, these experimental methods are not only expensive, but also have...

AI Technical Summary

Problems solved by technology

But so far there is no perfect method to predict the three-dimensional structure of protein

At present, the main technical bottleneck lies in two aspects: the first aspect is that the sampling ability of the existing technology to the conformational space is not strong; the other is that the energy function is not accurate, resulting in the conformation with the lowest energy not necessarily corresponding to the natural structure

[0004] Therefore, existing protein structure prediction methods have problems such as inaccurate energy functions, low sampling efficiency, and insufficient prediction accuracy, which need to be improved.

Images