Article
Biochemical Research Methods
Shuya Nakata, Yoshiharu Mori, Shigenori Tanaka
Summary: We propose an equivariant diffusion-based generative model that learns the joint distribution of ligand and protein conformations conditioned on the molecular graph of a ligand and the sequence representation of a protein extracted from a pre-trained protein language model. Benchmark results show that this protein structure-free model is capable of generating diverse structures of protein-ligand complexes, including those with correct binding poses. Further analyses indicate that the proposed end-to-end approach is particularly effective when the ligand-bound protein structure is not available.
BMC BIOINFORMATICS
(2023)
Article
Geriatrics & Gerontology
Xiaofeng Wang, Naixu Shi, Baiao Wu, Lin Yuan, Jiapeng Chen, Cong Ye, Miao Hao
Summary: Periodontitis is a high-risk factor for Parkinson's disease (PD), and the association between these two conditions is mainly manifested in immune and dopamine-related pathways. Hub genes, such as CDSN, TH, DDC, and SLC6A3, may serve as potential biomarkers for diagnosing or detecting PD.
FRONTIERS IN AGING NEUROSCIENCE
(2022)
Article
Computer Science, Artificial Intelligence
Yang Yu, Sez Atamturktur
Summary: In this study, a novel knowledge-informed generative adversarial network (KI-GAN) is proposed for the functional calibration of computer models under uncertainty. The proposed KI-GAN leverages the superior distribution learning ability of generative models for uncertainty quantification of model parameters and uses prior knowledge to inform generative model training. Two experiments are presented to demonstrate the effectiveness of the proposed KI-GAN for calibrating the functional model parameters under various scenarios.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Biochemical Research Methods
Jianmin Wang, Yanyi Chu, Jiashun Mao, Hyeon-Nae Jeon, Haiyan Jin, Amir Zeb, Yuil Jang, Kwang-Hwi Cho, Tao Song, Kyoung Tai No
Summary: This study constructs a dataset for protein-protein interaction (PPI) targeted drug-likeness and proposes a deep molecular generative framework to generate novel drug-like molecules based on the features of seed compounds. The results show that the generated molecules have better PPI-targeted drug-likeness and drug-likeness, and the model performs comparably to other state-of-the-art molecule generation models.
BRIEFINGS IN BIOINFORMATICS
(2022)
Article
Multidisciplinary Sciences
Hao Meng, Fei Yuan, Yang Tian, Hongwei Wei
Summary: This paper proposes a mixed attention model consisting of pixel attention model (PAM) and feature attention model (FAM) to improve the accuracy of ship detection in complex backgrounds. Experimental results show that using this method with Yolov3 increases mean average precision by 2.2%, achieving 0.975.
ARABIAN JOURNAL FOR SCIENCE AND ENGINEERING
(2022)
Article
Biochemistry & Molecular Biology
Su-Bin Yoon, Yu-Chien (Calvin) Ma, Akaash Venkat, Chun-Yu (Audi) Liu, Jie J. Zheng
Summary: This study investigated the genetic basis of Retinitis Pigmentosa (RP) and found that all causal genes of RP may belong within a complex network. By analyzing gene connections and protein interaction networks, the research successfully established network connections among RP genes and identified novel potential causal genes. The results suggest an interconnectedness causing RP at the molecular level.
INTERNATIONAL JOURNAL OF MOLECULAR SCIENCES
(2022)
Article
Genetics & Heredity
Zixuan Meng, Linai Kuang, Zhiping Chen, Zhen Zhang, Yihong Tan, Xueyong Li, Lei Wang
Summary: A prediction model called WPDINM is proposed in this study to detect key proteins based on a novel weighted protein-domain interaction network. Experimental results show that WPDINM achieves significantly higher predictive accuracy for key protein identification compared to traditional competing measures.
FRONTIERS IN GENETICS
(2021)
Article
Multidisciplinary Sciences
Amruta Sahoo, Sebastian Pechmann
Summary: This article defines functional network motifs (FNMs) through the integration of genetic interaction data and demonstrates their power in capturing regulatory interactions. It highlights the importance of FNMs for the systematic identification of feedback regulation in biological networks.
Article
Cell Biology
Shumei Zhang, Jingyu Zhang, Qichao Zhang, Yingjian Liang, Youwen Du, Guohua Wang
Summary: The study suggests that DNA methylation is associated with the occurrence of bladder cancer, and genes with differential DNA methylation levels can serve as potential biomarkers for the cancer. FASLG and PRKCZ have been identified as prognostic biomarkers for bladder cancer through Cox proportional hazard regression analysis. Patients can be categorized into high or low risk groups based on this two-gene prognostic model, which can help evaluate patient survival.
FRONTIERS IN CELL AND DEVELOPMENTAL BIOLOGY
(2022)
Article
Computer Science, Information Systems
Olorunjube James Falana, Adesina Simon Sodiya, Saidat Adebukola Onashoga, Biodun Surajudeen Badmus
Summary: Recent outbreaks of pandemics have led to an increase in cyberattacks caused by malware. This study proposes a novel ensemble technique, called Mal-Detect, which combines Deep Convolutional Neural Network and Deep Generative Adversarial Neural Network to analyze, detect, and categorize malware. Experimental results demonstrate that Mal-Detect outperforms other state-of-the-art techniques with an accuracy of 99.8% in detecting malware.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2022)
Article
Biochemistry & Molecular Biology
Apichat Suratanee, Kitiporn Plaimas
Summary: This study utilized a hybrid deep learning algorithm combined with a heterogeneous network to infer the functions of malaria parasite genes, identifying new gene functions that enhance our understanding of genome communications.
INTERNATIONAL JOURNAL OF MOLECULAR SCIENCES
(2021)
Article
Computer Science, Artificial Intelligence
Cong Tran, Won-Yong Shin, Andreas Spitz, Michael Gertz
Summary: In this paper, the authors propose a novel method called DeepNC for inferring the missing parts of a network based on a deep generative model. The method first learns the likelihood of edges and then identifies the graph that maximizes the learned likelihood conditioned on the observable graph topology. The authors empirically demonstrate the superiority of DeepNC over state-of-the-art network completion approaches.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2022)
Article
Multidisciplinary Sciences
Lucas Seninge, Ioannis Anastopoulos, Hongxu Ding, Joshua Stuart
Summary: The study introduces a novel sparse Variational Autoencoder architecture, VEGA, with gene module interpretability to provide deeper biological insights in transcriptomics data analysis. Through experiments in diverse biological contexts, VEGA successfully replicates cellular-specific response mechanisms, master regulator statuses, and reveals cell type and state identities in developing cells.
NATURE COMMUNICATIONS
(2021)
Article
Computer Science, Information Systems
Devika Kuttala, Dwarikanath Mahapatra, Ramanathan Subramanian, V. Ramana Murthy Oruganti
Summary: This study investigates Autism Spectrum Disorder (ASD) and Attention-deficit/hyperactivity disorder (ADHD), two neuro-developmental disorders in children. The researchers propose a framework based on a pediatrician's approach and develop a one-class model for characterizing healthy subjects. Using a Dense GAN architecture with self-attention modules, they train their framework using longitudinal data to better diagnose ASD and ADHD compared to existing methods.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2022)
Article
Computer Science, Information Systems
Angel Gonzalez-Prieto, Alberto Mozo, Sandra Gomez-Canaval, Edgar Talavera
Summary: This paper proposes a novel activation function based on the Smirnov probabilistic transformation to improve the quality of generated data in Generative Adversarial Networks (GANs). Unlike previous works, this activation function is applicable to any type of data distribution and can be seamlessly integrated into GAN training processes. Experimental results demonstrate that using this new activation function can significantly enhance the quality of generated data in GANs.
INFORMATION SCIENCES
(2022)
Article
Computer Science, Artificial Intelligence
Yu Wang, Wei Cui, Nhu Khue Vuong, Zhenghua Chen, Yu Zhou, Min Wu
Summary: Today's manufacturing systems are complex and dynamic. Predicting product quality is crucial for improving accuracy and productivity. Recent developments in artificial intelligence, especially machine learning, offer great potential for analyzing large amounts of manufacturing data. This study proposes an end-to-end framework using transfer learning for cross-machine product quality prediction, achieving significant improvement compared to conventional techniques.
JOURNAL OF INTELLIGENT MANUFACTURING
(2023)
Article
Computer Science, Information Systems
Zhongzhou Liu, Yuan Fang, Min Wu
Summary: Recommendation systems commonly overlook the popularity bias issue, which can impact the fairness of recommendations. To address this, the paper proposes a fairness-centric model, FAiR, that adaptively mitigates popularity bias for users and items. The model includes explicit fairness discriminators at the local level and an implicit discriminator at the global level, tailored to each individual user or item. Experimental results show that the model significantly outperforms state-of-the-art baselines in fairness metrics, while remaining competitive in effectiveness.
ACM TRANSACTIONS ON INFORMATION SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Xuefeng Su, Ru Li, Xiaoli Li, Baobao Chang, Zhiwei Hu, Xiaoqi Han, Zhichao Yan
Summary: Frame-semantic Parsing (FSP) is a challenging and critical task in NLP. Existing studies suffer from error propagation and inefficient solutions. To address these problems, we propose a novel target-aware relation classification task and a lightweight jointly learning framework. Experimental results demonstrate significant improvement compared to state-of-the-art models.
ACM TRANSACTIONS ON ASIAN AND LOW-RESOURCE LANGUAGE INFORMATION PROCESSING
(2023)
Article
Computer Science, Artificial Intelligence
Hongxiang Gao, Min Wu, Zhenghua Chen, Yuwen Li, Xingyao Wang, Shan An, Jianqing Li, Chengyu Liu
Summary: Facial expression recognition (FER) is a method of identifying emotions in facial photographs. Despite progress, three major challenges remain unaddressed: the interaction between spatial action units, inadequate semantic information about spectral expressions, and unbalanced data distribution. In this work, we propose SSA-ICL, a novel approach that solves these challenges by integrating spectral semantics with spatial locations and using an Intra-dataset Continual Learning (ICL) module to address data distribution.
Article
Automation & Control Systems
Qing Xu, Min Wu, Edwin Khoo, Zhenghua Chen, Xiaoli Li
Summary: In this paper, a hybrid deep learning model is proposed for early prediction of the remaining useful life (RUL) of lithium-ion batteries. This method effectively combines handcrafted features with latent features learned by deep networks to improve RUL prediction performance. A non-linear correlation-based method is also introduced to select effective domain knowledge-based features. A novel snapshot ensemble learning strategy is proposed to enhance model generalization ability. Experimental results demonstrate that the proposed method outperforms other approaches in both primary and secondary test sets.
IEEE-CAA JOURNAL OF AUTOMATICA SINICA
(2023)
Article
Computer Science, Artificial Intelligence
Emadeldeen Eldele, Mohamed Ragab, Zhenghua Chen, Min Wu, Chee-Keong Kwoh, Xiaoli Li, Cuntai Guan
Summary: Sleep staging is crucial for diagnosing and treating sleep disorders. Current data-driven deep learning models for automatic sleep staging have limitations when dealing with real-world scenarios. To overcome these limitations, this study proposes a novel adversarial learning framework called ADAST, which addresses the domain shift problem in the unlabeled target domain. The proposed framework outperforms state-of-the-art methods in six cross-domain scenarios.
IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTATIONAL INTELLIGENCE
(2023)
Article
Computer Science, Artificial Intelligence
Harikumar Kandath, Md Meftahul Ferdaus, Zhen Wei Ng, Bangjian Zhou, Suresh Sundaram, Xiaoli Li, Senthilnath Jayavelu
Summary: Researchers have developed intelligent Kalman filters (KFs) by combining them with machine learning algorithms for accurate estimation of states in dynamical systems. In this paper, a parsimonious autonomous sequential estimator (PASE) is proposed, which combines a KF-based estimator and an autonomous-structured recurrent parsimonious learning machine (rPALM) in a sequential manner. The performance evaluation shows that PASE provides better estimation accuracy with a compact architecture for both linear and nonlinear dynamical systems, making it suitable for real-world applications.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Zhenghua Chen, Min Wu, Alvin Chan, Xiaoli Li, Yew-Soon Ong
Summary: Artificial Intelligence (AI) is a rapidly growing field that promises significant benefits for consumers and businesses, but its development has come at a cost to the environment and raised concerns about its societal impacts. This review explores machine learning approaches to address the sustainability problem of AI and proposes research challenges and directions for the next generation of sustainable AI techniques.
IEEE COMPUTATIONAL INTELLIGENCE MAGAZINE
(2023)
Article
Engineering, Electrical & Electronic
Keyu Wu, Min Wu, Zhenghua Chen, Ruibing Jin, Wei Cui, Zhiguang Cao, Xiaoli Li
Summary: Domain adaptation transfers knowledge from label-rich source domains to label-scarce target domains for generalized learning in new environments. Partial domain adaptation (PDA) extends this concept by considering scenarios where the target label space is a subset of the source label space. This paper proposes a Reinforced Adaptation Network (RAN) that combines deep reinforcement learning with domain adaptation techniques to address the challenging PDA problem. Experimental results show that RAN significantly outperforms existing state-of-the-art methods on three benchmark datasets.
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
(2023)
Article
Computer Science, Hardware & Architecture
Devki Nandan Jha, Zhenghua Chen, Shudong Liu, Min Wu, Jiahan Zhang, Graham Morgan, Rajiv Ranjan, Xiaoli Li
Summary: Personalised health and fitness provide users with information regarding their wellbeing and an opportunity to inform healthcare services for better patient outcomes. AE-HAR is a model that balances battery depletion, data accuracy, and timely delivery of results in human activity recognition (HAR). It incorporates a lightweight machine learning component and cloud-based calculations to achieve high accuracy and energy consumption savings.
IEEE TRANSACTIONS ON SUSTAINABLE COMPUTING
(2023)
Article
Engineering, Biomedical
Emadeldeen Eldele, Mohamed Ragab, Zhenghua Chen, Min Wu, Chee-Keong Kwoh, Xiaoli Li
Summary: In the past few years, there has been a significant advancement in deep learning for EEG-based sleep stage classification. However, the success of these models relies on a large amount of labeled data for training, making them less applicable in real-world scenarios. Self-supervised learning has emerged as a successful technique to overcome the scarcity of labeled data. This paper evaluates the effectiveness of self-supervised learning in improving the performance of existing sleep stage classification models with limited labels.
IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Yong Liu, Susen Yang, Yonghui Xu, Chunyan Miao, Min Wu, Juyong Zhang
Summary: This paper proposes a recommendation framework called CGAT, which explicitly utilizes both local and non-local graph context information of entities in a knowledge graph. CGAT captures local context information using a user-specific graph attention mechanism, and extracts non-local context using biased random walk sampling process and models the dependency using an RNN. It also incorporates an item-specific attention mechanism to capture the user's personalized preferences. Experimental results on real datasets demonstrate the effectiveness of CGAT compared to state-of-the-art KG-based recommendation methods.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Yucheng Wang, Min Wu, Ruibing Jin, Xiaoli Li, Lihua Xie, Zhenghua Chen
Summary: Remaining useful life (RUL) prediction is crucial for prognostics and health management of a system. Deep learning models have emerged as leading solutions due to their powerful ability in nonlinear modeling and capturing temporal dependencies. However, existing methods have limitations in effectively modeling and capturing spatial dependencies. To overcome these limitations, we propose a novel framework that combines both local and global information to accurately predict RUL.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
