An Improved Prorank Algorithm for Detection of Overlapping Protein Complexes using Protein Interaction and Gene Expression Datasets in Yeast

Abstract

Detection of protein complex plays a significant role in holding out various biological purposes and it can be applied to learning the pathways from the network of Protein-Protein Interactions (PPI). The increasing amount of computational techniques for predicting the protein complex has been motivated by high-throughput experimental techniques that create a huge sum of protein interaction datasets. From the high-throughput technologies, PPI data are gathered and the datasets contain noisy. The clustering algorithms on the interaction datasets are usually not enough to attain trustworthy prediction results. The possible way is to enhance the accuracy of protein complex detection is achieved by exploiting the domain-domain interactions (DDI) and Protein-Protein Interactions (PPI). In this research, an improved prorank algorithm has been developed to detect the overlapping protein complexes. The rough set fuzzy algorithm is integrated with the prorank algorithm behind the ranking step to find the overlapping protein complexes. The proposed and the existing algorithms such as Improved Prorank, Prorank, CAMWI and PEWCC are examined on different datasets namely Protein-Protein Interaction (PPI) and Gene Expression such as GSE3076, GSE16799, GSE19213, PPI-D1, PPI-D2, Gavin and MIPS. The experimental outcomes illustrate that the proposed Improved prorank algorithm predicts more complex networks with superior results contrast to other techniques.