Identification of Protein Complexes Based on Core-Attachment Structure and Combination of Centrality Measures and Biolog
- PDF / 2,947,208 Bytes
- 22 Pages / 595.276 x 790.866 pts Page_size
- 27 Downloads / 203 Views
Identification of Protein Complexes Based on Core‑Attachment Structure and Combination of Centrality Measures and Biological Properties in PPI Weighted Networks Abdolkarim Elahi1 · Seyed Morteza Babamir1 Accepted: 28 September 2020 © Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract In protein interaction networks, a complex is a group of proteins that causes a biological process to take place. The correct identification of complexes can help to better understand function of cells used for therapeutic purposes, such as drug discoveries. This paper uses core-attachment structure, centrality measures, and biological properties of proteins to identify protein complex with the aim of enhancing prediction accuracy compared to related work. We used the inherent organization of complex to the identification in this article, while most methods have not considered such properties. On the other hand, clustering methods, as the common method for identifying complexes in protein interaction networks have been applied. However, we want to propose a method for more accurate identification of complexes in this article. Using this method, we determined the core center of each complex and its attachment proteins using the centrality measures, biological properties and weight density, whereby the weight of each interaction was calculated using the protein information in the gene ontology. In the proposed approach to weighting the network and measuring the importance of proteins, we used our previous work. To compare with other methods, we used datasets DIP, Collins, Krogan, and Human. The results show that the performance of our method was significantly improved, compared to other methods, in terms of detecting the protein complex. Using the p-value concept, we show the biological significance of our predicted complexes. The proposed method could identify an acceptable number of protein complexes, with the highest proportion of biological significance in collaborating on the functional annotation of proteins. Keywords Protein complex · Protein interaction network · Centrality measures · Essential protein · Core and attachment algorithm
1 Introduction Proteins are responsible for many activities and behaviors in the cell. Generally, a specific biological process is conducted during interaction with each other. With the advent of laboratory technologies with mass output, large protein interactions have been created. The protein–protein interaction (PPI) network is a network of physical interactions among proteins [1], in which the proteins are considered as vertices, and the physical interactions between the pairs of proteins are considered as graph edges. PPI networks play * Seyed Morteza Babamir [email protected] 1
Department of Computer, University of Kashan, Kashan, Iran
a very important role in biological processes, including discrimination, protein folding, signal transduction, transcription and translation [2]. The identification of protein complexes in the PPI network has greatly contribu
Data Loading...
