A method for identifying protein complexes based on BSO optimization

Abstract

The invention provides a method for identifying protein complexes based on BSO optimization. It includes the following steps: using the powerful global optimization ability of the BSO algorithm, the protein interaction network is regarded as a network-wide connectivity graph, and the distance between protein nodes is defined by combining the gene ontology annotation function information of the protein and the topology structure of the protein interaction network. The improved k-means algorithm performs preliminary clustering. Then, according to the four optimization principles of the BSO algorithm, a new fitness value is generated, and the in-module and out-of-module optimization operations are performed on the preliminarily formed protein modules, and the loop is iterated to find a better global solution, and finally the post-processing process is performed. The method of the invention can maintain the diversity of the population in the optimization process, thereby avoiding falling into local optimum, obtaining global optimum module division, and obtaining protein complexes with significant bioaccumulation.

Description

technical field [0001] The invention relates to the technical field of systems biology, in particular to a method for identifying protein complexes based on BSO optimization. Background technique [0002] The smooth progress of the Human Genome Project marks the arrival of the post-genome era. Protein is an essential component of all cell and tissue structures, and is the most important material basis for life activities, but each protein in a cell does not exist in isolation, but interacts with other proteins to perform its functions . Protein interaction plays a central role in life, not only the basis of normal physiological processes such as DNA replication, transcription, translation, metabolism, signal transduction, and cell cycle control, but also plays a crucial role in pathological processes. All protein interactions in organisms constitute a protein-protein interaction network (Protein-Protein Interaction Network, referred to as PPI network). Identifying protein...

AI Technical Summary

Problems solved by technology

[0003] At present, the study of disease-causing genes by accurately identifying protein complexes from the protein interaction network (PPI) is a hot spot in biomedical research at home and abroad. However, due to the large number of proteins and their interactions, and the lack of There are a large number of proteins and their interaction functions that have not yet been labeled. Using biological experimental methods to test and analyze one by one is time-consuming, expensive and unsustainable.

Images