K-means Clustering: An Efficient Algorithm for Protein Complex Detection

The protein complexes have significant biological functions of proteins and nucleic acids dense from the molecular interaction network in cells. Several computational methods are developed to detect protein complexes from the protein protein interaction (PPI) networks. The existing algorithms do not predict better complex, and it also provides low performance values. In this research, K-means algorithm has been proposed for protein complex detection and compared with the existing algorithms such as MCODE and SPICi. The protein interaction and gene expression benchmark datasets such as Collins, DIP, Krogan, Krogan Extended, PPI-D1, PPI-D2, GSE12220, GSE12221, GSE12442, and GSE17716 have been used for comparing the performance of the existing and proposed algorithms. From this experimental analysis, it is inferred that the proposed K-means clustering algorithm outperforms the other existing methods.