Predicting multifunctional peptides based on a multi-scale ResNet model combined with channel attention mechanisms

Jing Liu; Hongpu Zhao; Yu Zhang; Jin Liu; Xiao Guan; Jing Liu; Hongpu Zhao; Yu Zhang; Jin Liu; Xiao Guan

doi:10.3934/era.2024133

Electronic Research Archive

2024, Volume 32, Issue 4: 2921-2935. doi: 10.3934/era.2024133

Previous Article Next Article

Research article

Predicting multifunctional peptides based on a multi-scale ResNet model combined with channel attention mechanisms

1.
College of Information Engineering, Shanghai Maritime University, Shanghai 201306, China
2.
School of Health Science and Engineering, University of Shanghai for Science and Technology, Shanghai 200093, China
3.
National Grain Industry (Urban Grain and Oil Security) Technology Innovation Center, Shanghai 200093, China

Received: 03 February 2024 Revised: 09 March 2024 Accepted: 02 April 2024 Published: 16 April 2024

Peptides are biomolecules composed of multiple amino acid residues connected by peptide bonds, which are widely involved in physiological and biochemical processes in organisms and exhibit diverse functions. In previous studies, the focus was primarily on single-functional peptides. However, research trends indicate that an increasing number of multifunctional peptides are being identified and discovered. To address this challenge, we proposed a deep learning method based on multi-scale ResNet as the backbone combined with a channel attention mechanism (called MSRC) for the identification of multifunctional peptides. Furthermore, the data imbalance problem was solved through the comprehensive use of online data augmentation and confidence-based weighted loss functions. Experimental results demonstrated that the proposed MSRC method achieved an accuracy of 0.688 with an absolute true rate of 0.619. Notably, in predicting minority class peptides such as AEP, AHIVP, and BBP, the MSRC model exhibited heightened sensitivity, showcasing its exceptional capability in addressing issues related to minority classes. By enhancing the precision in identifying and predicting multifunctional peptides, the MSRC method was poised to contribute significantly to advancements in drug discovery, disease treatment, and biotechnology.
- multifunctional peptides,
- multi-scale ResNet,
- channel attention,
- data augmentation,
- optimization loss
Citation: Jing Liu, Hongpu Zhao, Yu Zhang, Jin Liu, Xiao Guan. Predicting multifunctional peptides based on a multi-scale ResNet model combined with channel attention mechanisms[J]. Electronic Research Archive, 2024, 32(4): 2921-2935. doi: 10.3934/era.2024133

Related Papers:

Abstract

Peptides are biomolecules composed of multiple amino acid residues connected by peptide bonds, which are widely involved in physiological and biochemical processes in organisms and exhibit diverse functions. In previous studies, the focus was primarily on single-functional peptides. However, research trends indicate that an increasing number of multifunctional peptides are being identified and discovered. To address this challenge, we proposed a deep learning method based on multi-scale ResNet as the backbone combined with a channel attention mechanism (called MSRC) for the identification of multifunctional peptides. Furthermore, the data imbalance problem was solved through the comprehensive use of online data augmentation and confidence-based weighted loss functions. Experimental results demonstrated that the proposed MSRC method achieved an accuracy of 0.688 with an absolute true rate of 0.619. Notably, in predicting minority class peptides such as AEP, AHIVP, and BBP, the MSRC model exhibited heightened sensitivity, showcasing its exceptional capability in addressing issues related to minority classes. By enhancing the precision in identifying and predicting multifunctional peptides, the MSRC method was poised to contribute significantly to advancements in drug discovery, disease treatment, and biotechnology.

References

[1]	C. Guntuboina, A. Das, P. Mollaei, S. Kim, A. B. Farimani, Peptidebert: A language model based on transformers for peptide property prediction, J. Phys. Chem. Lett., 14 (2023), 10427–10434. https://doi.org/10.1021/acs.jpclett.3c02398 doi: 10.1021/acs.jpclett.3c02398
[2]	M. Muttenthaler, G. F. King, D. J. Adams, P. F. Alewood, Trends in peptide drug discovery, Nat. Rev. Drug Discovery, 20 (2021), 309–325. https://doi.org/10.1038/s41573-020-00135-8 doi: 10.1038/s41573-020-00135-8
[3]	E. B. M. Daliri, B. H. Lee, D. H. Oh, Current trends and perspectives of bioactive peptides, Crit. Rev. Food Sci. Nutr., 58 (2018), 2273–2284. https://doi.org/10.1080/10408398.2017.1319795 doi: 10.1080/10408398.2017.1319795
[4]	W. Tang, R. Dai, W. Yan, W. Zhang, Y. Bin, E. Xia, et al., Identifying multi-functional bioactive peptide functions using multi-label deep learning, Briefings Bioinf., 23 (2022), bbab414. https://doi.org/10.1093/bib/bbab414 doi: 10.1093/bib/bbab414
[5]	Y. Ma, Z. Guo, B. Xia, Y. Zhang, X. Liu, Y. Yu, et al., Identification of antimicrobial peptides from the human gut microbiome using deep learning, Nat. Biotechnol., 40 (2022), 921–931. https://doi.org/10.1038/s41587-022-01226-0 doi: 10.1038/s41587-022-01226-0
[6]	Y. Ma, X. Liu, X. Zhang, Y. Yu, Y. Li, M. Song, et al., Efficient mining of anticancer peptides from gut metagenome, Adv. Sci., 10 (2023), 2300107. https://doi.org/10.1002/advs.202300107 doi: 10.1002/advs.202300107
[7]	J. Zhang, Z. Zhang, L. Pu, J. Tang, F. Guo, AIEpred: An ensemble predictive model of classifier chain to identify anti-inflammatory peptides, IEEE/ACM Trans. Comput. Biol. Bioinf., 18 (2020), 1831–1840. https://doi.org/10.1109/TCBB.2020.2968419 doi: 10.1109/TCBB.2020.2968419
[8]	F. F. Atanaki, S. Behrouzi, S. Ariaeenejad, A. Boroomand, K. Kavousi, BIPEP: Sequence-based prediction of biofilm inhibitory peptides using a combination of NMR and physicochemical descriptors, ACS Omega, 5 (2020), 7290–7297. https://doi.org/10.1021/acsomega.9b04119 doi: 10.1021/acsomega.9b04119
[9]	K. Liu, Y. Fu, L. Wu, X. Li, C. Aggarwal, H. Xiong, Automated feature selection: A reinforcement learning perspective, IEEE Trans. Knowl. Data Eng., 35 (2023), 2272–2284. https://doi.org/10.1109/TKDE.2021.3115477 doi: 10.1109/TKDE.2021.3115477
[10]	P. Agrawal, D. Bhagat, M. Mahalwal, N. Sharma, G. P. S. Raghava, AntiCP 2.0: An updated model for predicting anticancer peptides, Briefings Bioinf., 22 (2021), bbaa153. https://doi.org/10.1093/bib/bbaa153 doi: 10.1093/bib/bbaa153
[11]	W. Zhang, E. Xia, R. Dai, W. Tang, Y. Bin, J. Xia, PredAPP: Predicting anti-parasitic peptides with undersampling and ensemble approaches, Interdiscip. Sci.: Comput. Life Sci., 14 (2022), 258–268. https://doi.org/10.1007/s12539-021-00484-x doi: 10.1007/s12539-021-00484-x
[12]	B. Manavalan, T. H. Shin, M. O. Kim, G. Lee, AIPpred: Sequence-based prediction of anti-inflammatory peptides using random forest, Front. Pharmacol., 9 (2018), 348997. https://doi.org/10.3389/fphar.2018.00276 doi: 10.3389/fphar.2018.00276
[13]	Y. Han, D. Kim, Deep convolutional neural networks for pan-specific peptide-MHC class I binding prediction, BMC Bioinf., 18 (2017), 585. https://doi.org/10.1186/s12859-017-1997-x doi: 10.1186/s12859-017-1997-x
[14]	Y. Hu, Z. Wang, H. Hu, F. Wan, L. Chen, Y. Xiong, et al., ACME: Pan-specific peptide–MHC class I binding prediction through attention-based deep neural networks, Bioinformatics, 35 (2019), 4946–4954. https://doi.org/10.1093/bioinformatics/btz427 doi: 10.1093/bioinformatics/btz427
[15]	H. C. Yi, Z. H. You, X. Zhou, L. Cheng, X. Li, T. Jiang, et al., ACP-DL: A deep learning long short-term memory model to predict anticancer peptides using high-efficiency feature representation, Mol. Ther. Nucleic Acids, 17 (2019), 1–9. https://doi.org/10.1016/j.omtn.2019.04.025 doi: 10.1016/j.omtn.2019.04.025
[16]	A. Ghulam, F. Ali, R. Sikander, A. Ahmad, A. Ahmed, S. Patil, ACP-2DCNN: Deep learning-based model for improving prediction of anticancer peptides using two-dimensional convolutional neural network, Chemom. Intell. Lab. Syst., 226 (2022), 104589. https://doi.org/10.1016/j.chemolab.2022.104589 doi: 10.1016/j.chemolab.2022.104589
[17]	L. Yu, R. Jing, F. Liu, J. Luo, Y. Li, DeepACP: A novel computational approach for accurate identification of anticancer peptides by deep learning algorithm, Mol. Ther. Nucleic Acids, 22 (2020), 862–870. https://doi.org/10.1016/j.omtn.2020.10.005 doi: 10.1016/j.omtn.2020.10.005
[18]	J. M. Conlon, M. Mechkarska, M. L. Lukic, P. R. Flatt, Potential therapeutic applications of multifunctional host-defense peptides from frog skin as anti-cancer, anti-viral, immunomodulatory, and anti-diabetic agents, Peptides, 57 (2014), 67–77. https://doi.org/10.1016/j.peptides.2014.04.019
[19]	H. Fan, W. Yan, L. Wang, J. Liu, Y. Bin, J. Xia, Deep learning-based multi-functional therapeutic peptides prediction with a multi-label focal dice loss function, Bioinformatics, 39 (2023), btad334. https://doi.org/10.1093/bioinformatics/btad334 doi: 10.1093/bioinformatics/btad334
[20]	H. Lv, K. Yan, B. Liu, TPpred-LE: Therapeutic peptide function prediction based on label embedding, BMC Biol., 21 (2023), 238. https://doi.org/10.1186/s12915-023-01740-w doi: 10.1186/s12915-023-01740-w
[21]	Y. Li, X. Li, Y. Liu, Y. Yao, G. Huang, MPMABP: A CNN and Bi-LSTM-Based method for predicting multi-activities of bioactive peptides, Pharmaceuticals, 15 (2022), 707. https://doi.org/10.3390/ph15060707 doi: 10.3390/ph15060707
[22]	W. Lin, D. Xu, Imbalanced multi-label learning for identifying antimicrobial peptides and their functional types, Bioinformatics, 32 (2016), 3745–3752. https://doi.org/10.1093/bioinformatics/btw560 doi: 10.1093/bioinformatics/btw560
[23]	W. Yan, W. Tang, L. Wang, Y. Bin, J. Xia, PrMFTP: Multi-functional therapeutic peptides prediction based on multi-head self-attention mechanism and class weight optimization, PLoS Comput. Biol., 18 (2022), e1010511. https://doi.org/10.1371/journal.pcbi.1010511 doi: 10.1371/journal.pcbi.1010511
[24]	H. Kim, J. H. Jang, S. C. Kim, J. H. Cho, De novo generation of short antimicrobial peptides with enhanced stability and cell specificity, J. Antimicrob. Chemother., 69 (2014), 121–132. https://doi.org/10.1093/jac/dkt322 doi: 10.1093/jac/dkt322
[25]	E. Vušak, V. Kužina, A. Jović, A survey of word embedding algorithms for textual data information extraction, in 2021 44th International Convention on Information, Communication and Electronic Technology (MIPRO), IEEE, (2021), 181–186. https://ieeexplore.ieee.org/document/9597076
[26]	F. Ge, Y. Zhang, J. Xu, A. Muhammad, J. Song, D. Yu, Prediction of disease-associated nsSNPs by integrating multi-scale ResNet models with deep feature fusion, Briefings Bioinf., 23 (2022), bbab530. https://doi.org/10.1093/bib/bbab530 doi: 10.1093/bib/bbab530
[27]	Z. Zhao, J. Gui, A. Yao, N. Q. K. Le, M. C. H. Chua, Improved prediction model of protein and peptide toxicity by integrating channel attention into a convolutional neural network and gated recurrent units, ACS Omega, 7 (2022), 40569–40577. https://doi.org/10.1021/acsomega.2c05881 doi: 10.1021/acsomega.2c05881
[28]	T. Zhu, X. Liu, E. Zhu, Oversampling with reliably expanding minority class regions for imbalanced data learning, IEEE Trans. Knowl. Data Eng., 35 (2023), 6167–6181. https://ieeexplore.ieee.org/document/9773030
[29]	D. Wang, H. Yu, G. Fan, Facial action unit recognition algorithm based on deep learning (in Chinese), J. East China Univ. Sci. Technol. (Nat. Sci. Ed.), 46 (2020), 269–276. https://doi.org/10.14135/j.cnki.1006-3080.20190107003 doi: 10.14135/j.cnki.1006-3080.20190107003
[30]	A. Elnaggar, M. Heinzinger, C. Dallago, G. Rihawi, Y. Wang, L. Jones, et al., ProtTrans: Towards cracking the language of life's code through self-supervised deep learning and high performance computing, preprint, arXiv: 2007.06225.
[31]	Y. Zhang, G. Zhu, K. Li, F. Li, L. Huang, M. Duan, et al., HLAB: Learning the BiLSTM features from the ProtBert-encoded proteins for the class I HLA-peptide binding prediction, Briefings Bioinf., 23 (2022), bbac173. https://doi.org/10.1093/bib/bbac173 doi: 10.1093/bib/bbac173

Reader Comments

Your name:*

Email:*
© 2024 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)