Prediction method and prediction device for protein side chain

Abstract

The invention discloses a prediction method and prediction device for a protein side chain. The prediction method comprises the first step of selecting the number of target ant colonies and initializing ant colony algorithm parameters; the second step of initializing protein side chain conformations and establishing a mapping relation between the protein side chain conformations and a side chain rotational isomer library; the third step of selecting the optimizing sequence of protein residues randomly through the target ant colonies and reconstructing the protein side chain in the target ant colonies with the guide of a pheromone matrix and an enlightenment value, wherein if circulation is carried out even-number times, local optimization is carried out continuously, energy calculation is carried out on protein of the reconstructed side chain in the target ant colonies through an energy function, a protein conformation with the minimum contemporary energy is selected and is compared with a best protein conformation, the best protein conformation is updated by selecting the protein with small energy, and the pheromone matrix is updated; the fourth step of executing the third step repeatedly until the end condition is met to obtain the optimal protein side chain conformation.

Description

technical field [0001] The invention relates to a protein-related data processing method, in particular to a protein side chain prediction method and a prediction device. Background technique [0002] Proteins are executors of specific life functions because of their specific structures. In this sense, proteins are the most important class of biological macromolecules. Therefore, determining the three-dimensional structure of proteins has become an important topic in the field of biology. With the successful completion of the Human Genome Project, the complete determination of the gene sequences of various model animals and plants, and the continuous development of protein engineering technology, the prediction of the three-dimensional structure of proteins has become more and more important. This is reflected in two aspects. One is that the biological information contained in the massive nucleic acid and protein primary structure databases urgently needs to be interpreted...

AI Technical Summary

Problems solved by technology

However, the sidechain prediction problem is an NP-C problem, so in theory, the deterministic algorithm cannot obtain the optimal solution

However, these algorithms all have a common shortcoming. They all use different energy functions, and their energy functions need to be trained, which means that these energy functions are not universal, and the side chain predictions for some proteins are more accurate.

Images