Article
Biochemical Research Methods
Xuan Xiao, Yu-Tao Shao, Zhen-Tao Luo, Wang-Ren Qiu
Summary: This paper aims to identify 5-methylcytosine sites in human promoters and constructs a predictor called m5C-HPromoter. The results demonstrate that m5C-HPromoter has good performance in terms of accuracy and sensitivity, and shows improvement compared to existing predictors.
CURRENT BIOINFORMATICS
(2022)
Article
Computer Science, Information Systems
Chen Wang, Junyin Zhang, Li Cheng, Jiawei Wu, Minfeng Xiao, Junfeng Xia, Yannan Bin
Summary: In this study, a two-layer model called DPProm is introduced for predicting phage promoters and their types. The first layer, DPProm-1L, uses a dual-channel deep neural network ensemble method to identify whether a DNA sequence is a promoter or non-promoter. The second layer, DPProm-2L, predicts the types of promoters. Experimental results show that DPProm outperforms existing methods and reduces false positive rate effectively. A user-friendly web interface is also provided.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2022)
Article
Computer Science, Information Systems
Yinuo Lyu, Wenying He, Shuhao Li, Quan Zou, Fei Guo
Summary: Promoters, regulatory elements located near transcription start sites, initiate gene transcription. A novel two-layer predictor, iPro2L-PSTKNC, based on a new feature extraction model, PSTKNC, is developed to identify E.coli genome promoters effectively. The ensemble classification SVM shows the best performance with high accuracy and MCC.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2021)
Article
Biochemical Research Methods
Fei He, Jingyi Li, Rui Wang, Xiaowei Zhao, Ye Han
Summary: This study proposed a novel deep learning architecture for predicting protein Ubiquitylation and SUMOylation sites as well as their crosstalk sites simultaneously. The method achieved promising AUCs of 0.838, 0.888, and 0.862 on Ubiquitylation, SUMOylation, and crosstalk sites respectively in tenfold cross-validation. The results also demonstrated the effectiveness of the proposed architecture in classifying Ubiquitylated and SUMOylated lysine residues.
BMC BIOINFORMATICS
(2021)
Article
Automation & Control Systems
Shanjiang Tang, Chunjiang Wang, Jiangtian Nie, Neeraj Kumar, Yang Zhang, Zehui Xiong, Ahmed Barnawi
Summary: Efficient screening of COVID-19 cases is crucial to prevent the rapid spread of the disease, and the EDL-COVID model, combining deep learning and ensemble learning, shows promising results in COVID-19 case detection with a higher accuracy compared to the COVID-Net model.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2021)
Article
Biochemical Research Methods
Yuanying Zhuang, Xiangrong Liu, Yue Zhong, Longxin Wu
Summary: This study proposes a comprehensive feature representation algorithm based on a deep ensemble model and convolutional neural network to identify anti-hypertension peptides. The results show that the performance of this method is better than other methods, providing important insights for hypertension therapy.
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2022)
Article
Biochemical Research Methods
Jun Zhang, Qingcai Chen, Bin Liu
Summary: DBPs and RBPs are crucial proteins associated with various cell activities and diseases. DeepDRBP-2L, combining CNN and LSTM, is the first computational method able to identify DBPs, RBPs and DRBPs, overcoming existing methods' shortcomings with high prediction accuracy.
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2021)
Article
Mathematical & Computational Biology
Shengming Zhou, Jia Zheng, Cangzhi Jia
Summary: Regulatory elements in DNA sequences are crucial for gene expression, with promoters being key in transcriptional regulation. The SPREAD model proposed in this study significantly improves promoter prediction performance in Pseudomonas aeruginosa.
MATHEMATICAL BIOSCIENCES AND ENGINEERING
(2022)
Article
Biochemical Research Methods
Yan Zhu, Fuyi Li, Xudong Guo, Xiaoyu Wang, Lachlan J. M. Coin, Geoffrey Webb, Jiangning Song, Cangzhi Jia
Summary: In this study, we developed TIMER, a Siamese neural network-based approach for identifying both general and species-specific bacterial promoters. TIMER achieves a competitive performance and outperforms several existing methods on both general and species-specific promoter prediction.
BRIEFINGS IN BIOINFORMATICS
(2023)
Article
Mathematics
Yongbing Chen, Wenyuan Qin, Tong Liu, Ruikun Li, Fei He, Ye Han, Zhiqiang Ma, Zilin Ren
Summary: N-terminal acetylation is a specific protein modification that plays a significant role in protein stability, folding, subcellular localization, and protein-protein interactions. We have developed MTNA, a deep learning network, which accurately predicts N-terminal protein acetylation sites for various amino acids at the N-terminus.
ELECTRONIC RESEARCH ARCHIVE
(2023)
Article
Computer Science, Interdisciplinary Applications
Huijuan Qiao, Shengli Zhang, Tian Xue, Jinyue Wang, Bowei Wang
Summary: In this study, a deep learning-based model called iPro-GAN is proposed for the identification of gene promoters and their strength. By using Moran-based spatial auto-cross correlation method as the feature extraction method and deep convolution generative adversarial network for classification, high accuracy is achieved. The results of the model are far superior to the existing best predictor, demonstrating its practicality.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2022)
Article
Computer Science, Interdisciplinary Applications
Zhi-min Zhang, Jian-ping Zhao, Pi-Jing Wei, Chun-Hou Zheng
Summary: The study introduced a new hybrid model to identify promoters and predict their strength, achieving higher accuracy by combining different neural networks and mechanisms.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2022)
Article
Computer Science, Artificial Intelligence
Moinul Islam, Md Tanzim Reza, Mohammed Kaosar, Mohammad Zavid Parvez
Summary: This study applies the Federated Learning (FL) approach to classify brain tumors from MRI images while preserving patient privacy. By training multiple CNN models and combining them into an ensemble classifier, the FL model achieves a slightly lower accuracy compared to the base ensemble model. However, the FL approach successfully protects patient privacy and demonstrates scalability.
NEURAL PROCESSING LETTERS
(2023)
Article
Biochemical Research Methods
Vishakha Singh, Sameer Shrivastava, Sanjay Kumar Singh, Abhinav Kumar, Sonal Saxena
Summary: Due to the emergence of multi-drug resistant bacteria, researchers are seeking alternatives to existing antibiotics in the form of antibacterial peptides (ABPs). In this study, a deep learning classifier based on a stacked ensemble technique was developed to classify peptides as antibacterial or not. A web app was also developed to identify novel ABPs in protein sequences. The model outperformed existing classifiers and showed potential for exploring new broad-spectrum ABPs.
BRIEFINGS IN BIOINFORMATICS
(2022)
Article
Biology
Fei Li, Shuai Liu, Kewei Li, Yaqi Zhang, Meiyu Duan, Zhaomin Yao, Gancheng Zhu, Yutong Guo, Ying Wang, Lan Huang, Fengfeng Zhou
Summary: DNA methylation is a major epigenetic modification that regulates biological processes without altering the DNA sequence. This study proposes a feature representation framework called EpiTEAmDNA, which integrates convolutional neural network and conventional machine learning methods. It shows improved performances compared to existing deep learning methods on small datasets across multiple DNA methylation types.
COMPUTERS IN BIOLOGY AND MEDICINE
(2023)