MM-CCNB: Essential protein prediction using MAX-MIN strategies and compartment of common neighboring approach. (January 2023)

• Record Type: Journal Article
• Authors: Payra, Anjan Kumar; Saha, Banani; Ghosh, Anupam
• Is Part Of: Computer methods and programs in biomedicine. Volume 228(2023)
• Publication Date: 2023-01
• Subjects: Common neighbor -- Edge clustering coefficient -- GO-attribute -- Subcellular localization -- Essentiality Score; Medicine -- Computer programs -- Periodicals; Biology -- Computer programs -- Periodicals; Computers -- Periodicals; Medicine -- Periodicals; Médecine -- Logiciels -- Périodiques; Biologie -- Logiciels -- Périodiques; Biology -- Computer programs; Medicine -- Computer programs; Periodicals; Electronic journals; 610.28
• DOI: 10.1016/j.cmpb.2022.107247
• Languages: English
• ISSNs: 0169-2607

Abstract

Highlights: The novelty of the algorithm is stated below: Pair wise and protein wise set of essential proteins are obtained using Modified Jaccard's Similarity co-efficient algorithm. Modified Jaccard's Similarity co-efficient predicts functionally enriched set (GO-Attribute) of essential proteins as shown in Table 9 and 10 . Functionally and location wise active set of common neighbor proteins have been identified using MM-CCNB algorithm. MAX-MIN game strategy is applied on feature essentiality score to find protein wise essentiality score. Due to MAX-MIN strategy, the prediction accuracy has been improved compared to other existing methodologies. Abstract: Background and objective: Proteins are indispensable for the flow of the life of living organisms. Protein pairs in interaction exhibit more functional activities than individuals. These activities have been considered an essential measure in predicting their essentiality. Neighborhood approaches have been used frequently in the prediction of essentiality scores. All paired neighbors of the essential proteins are nominated for the suitable candidate seeds for prediction. Still now Jaccard's coefficient is limited to predicting functions, homologous groups, sequence analysis, etc. It really motivate us to predict essential proteins efficiently using different computational approaches. Methods: In our work, we proposed modified Jaccard's coefficient to predict essential proteins. We have proposed a novel methodology for predicting essential proteins using MAX-MIN strategies and modified Jaccard's coefficient approach. Results: The performance of our proposed methodology has been analyzed for Saccharomyces cerevisiae datasets with an accuracy of more than 80%. It has been observed that the proposed algorithm is outperforms with an accuracy of 0.78, 0.74, 0.79, and 0.862 for YDIP, YMIPS, YHQ, and YMBD datasets respectivly. Conclusions: There are several computational approaches in the existing state-of-art model of essential protein prediction. It has been noted that our predicted methodology outperforms other existing models viz. different centralities, local interaction density combined with protein complexes, modified monkey algorithm and ortho_sim_loc methods.