Predicting protein-peptide binding sites with a deep convolutional neural network

JOURNAL OF THEORETICAL BIOLOGY

卷:496

Article

Wardah, Wafaa¹  Dehzangi, Abdollah²  Taherzadeh, Ghazaleh³  Rashid, Mahmood A.^4,5  Khan, M. G. M.¹  Tsunoda, Tatsuhiko^6,7,8,9  Sharma, Alok^5,7,8,10 

[1] Univ South Pacific, Fac Sci, Sch Comp Informat & Math Sci, Suva, Fiji

[2] Morgan State Univ, Dept Comp Sci, Baltimore, MD 21239 USA

[3] Univ Maryland, Inst Biosci & Biotechnol Res, College Pk, MD USA

[4] Victoria Univ, Inst Sustainable Ind & Liveable Cities, Melbourne, Vic, Australia

[5] Griffith Univ, Inst Integrated & Intelligent Syst, Nathan, Qld, Australia

[6] Tokyo Med & Dent Univ, Med Res Inst, Dept Med Sci Math, Tokyo, Japan

[7] RIKEN, Lab Med Sci Math, Ctr Integrat Med Sci, Yokohama, Kanagawa, Japan

[8] JST, CREST, Tokyo 1138510, Japan

[9] Univ Tokyo, Grad Sch Sci, Dept Biol Sci, Lab Med Sci Math, Tokyo, Japan

[10] Univ South Pacific, Sch Engn & Phys, Suva, Fiji

关键词: Protein-peptide binding;    Artificial intelligence;    Deep learning;    Convolutional neural network;    Protein sequence;   

DOI  :  10.1016/j.jtbi.2020.110278

来源: Elsevier

【 摘 要 】

Motivation: Interactions between proteins and peptides influence biological functions. Predicting such bio-molecular interactions can lead to faster disease prevention and help in drug discovery. Experimental methods for determining protein-peptide binding sites are costly and time-consuming. Therefore, computational methods have become prevalent. However, existing models show extremely low detection rates of actual peptide binding sites in proteins. To address this problem, we employed a two-stage technique first, we extracted the relevant features from protein sequences and transformed them into images applying a novel method and then, we applied a convolutional neural network to identify the peptide binding sites in proteins. Results: We found that our approach achieves 67% sensitivity or recall (true positive rate) surpassing existing methods by over 35%. (C) 2020 Elsevier Ltd. All rights reserved.

【 授权许可】

Free