A5100340964- Bioinformatics and Genomic Networks
- Computational Drug Discovery Methods
- Cancer Research and Treatments
- Domain Adaptation and Few-Shot Learning
- Anomaly Detection Techniques and Applications
- EEG and Brain-Computer Interfaces
- Cancer-related gene regulation
- Cancer Cells and Metastasis
- Adversarial Robustness in Machine Learning
- Time Series Analysis and Forecasting
- Gene expression and cancer classification
- Advanced Graph Neural Networks
- Advanced Neural Network Applications
- Neural Networks and Applications
- Functional Brain Connectivity Studies
- Epigenetics and DNA Methylation
- RNA modifications and cancer
- Cancer-related Molecular Pathways
- Fibroblast Growth Factor Research
- Natural Language Processing Techniques
- Monoclonal and Polyclonal Antibodies Research
- Cancer Genomics and Diagnostics
- Topic Modeling
- Multimodal Machine Learning Applications
- Machine Learning in Bioinformatics
Institute for Infocomm Research
2016-2025
Agency for Science, Technology and Research
2016-2025
Second Xiangya Hospital of Central South University
2023-2025
State Grid Corporation of China (China)
2022-2025
University of Hong Kong - Shenzhen Hospital
2024-2025
University of Hong Kong
2024-2025
Central South University
2008-2025
Yangzhou University
2023-2025
Sichuan University
2021-2025
Wuhan University
2021-2025
Microphthalmia (Mi) is a bHLHZip transcription factor that essential for melanocyte development and postnatal function. It thought to regulate both differentiated features of melanocytes such as pigmentation well proliferation/survival, based on phenotypes mutant mouse alleles. Mi activity controlled by at least two signaling pathways. Melanocyte-stimulating hormone (MSH) promotes the gene through cAMP elevation, resulting in sustained up-regulation over many hours. c-Kit up-regulates...
For prognostics and health management of mechanical systems, a core task is to predict the machine remaining useful life (RUL). Currently, deep structures with automatic feature learning, such as long short-term memory (LSTM), have achieved great performances for RUL prediction. However, conventional LSTM network only uses learned features at last time step regression or classification, which not efficient. Besides, some handcrafted domain knowledge may convey additional information...
Automatic sleep stage classification is of great importance to measure quality.In this paper, we propose a novel attention-based deep learning architecture called AttnSleep classify stages using single channel EEG signals.This starts with the feature extraction module based on multi-resolution convolutional neural network (MRCNN) and adaptive recalibration (AFR).The MRCNN can extract low high frequency features AFR able improve quality extracted by modeling inter-dependencies between...
Abstract Background How to detect protein complexes is an important and challenging task in post genomic era. As the increasing amount of protein-protein interaction (PPI) data are available, we able identify from PPI networks. However, most current studies based solely on observation that dense regions networks may correspond complexes, but fail consider inherent organization within complexes. Results To provide insights into this paper presents a novel core-attachment method (COACH) which...
In pharmaceutical sciences, a crucial step of the drug discovery process is identification drug-target interactions. However, only small portion interactions have been experimentally validated, as experimental validation laborious and costly. To improve efficiency, there great need for development accurate computational approaches that can predict potential to direct verification. this paper, we propose novel interaction prediction algorithm, namely neighborhood regularized logistic matrix...
Most proteins form macromolecular complexes to perform their biological functions. However, experimentally determined protein complex data, especially of those involving more than two partners, are relatively limited in the current state-of-the-art high-throughput experimental techniques. Nevertheless, many techniques (such as yeast-two-hybrid) have enabled systematic screening pairwise protein-protein interactions en masse. Thus computational approaches for detecting from interaction data...
Concolic testing combines program execution and symbolic analysis to explore the paths of a software program. In this paper, we develop first concolic approach for Deep Neural Networks (DNNs). More specifically, utilise quantified linear arithmetic over rationals express test requirements that have been studied in literature, then coherent method perform with aim better coverage. Our experimental results show effectiveness both achieving high coverage finding adversarial examples.
Learning decent representations from unlabeled time-series data with temporal dynamics is a very challenging task. In this paper, we propose an unsupervised Time-Series representation learning framework via Temporal and Contextual Contrasting (TS-TCC), to learn data. First, the raw are transformed into two different yet correlated views by using weak strong augmentations. Second, novel contrasting module robust designing tough cross-view prediction Last, further discriminative...
Experimental determination of drug-target interactions is expensive and time-consuming. Therefore, there a continuous demand for more accurate predictions using computational techniques. Algorithms have been devised to infer novel on global scale where the input these algorithms network (i.e., bipartite graph edges connect pairs drugs targets that are known interact). However, had difficulty predicting involving <italic xmlns:mml="http://www.w3.org/1998/Math/MathML"...
As one step toward the future smart grid, condition monitoring is important to facilitate reliability of grid asset operation and save on maintenance cost [1]. Most failures power are caused by electrical insulation failure, a key indicator such failure occurrence partial discharge (PD). Therefore, focus detect PD, especially in early stages, prevent serious or outage.
Abstract Spatial transcriptomics technologies generate gene expression profiles with spatial context, requiring spatially informed analysis tools for three key tasks, clustering, multisample integration, and cell-type deconvolution. We present GraphST, a graph self-supervised contrastive learning method that fully exploits data to outperform existing methods. It combines neural networks learn informative discriminative spot representations by minimizing the embedding distance between...
Recognition of human activities is an important task due to its far-reaching applications such as healthcare system, context-aware applications, and security monitoring. Recently, WiFi based activity recognition (HAR) becoming ubiquitous non-invasiveness. Existing WiFi-based HAR methods regard signals a temporal sequence channel state information (CSI), employ deep sequential models (e.g., RNN, LSTM) automatically capture channel-over-time features. Although being remarkably effective, they...
The remaining useful life (RUL) prediction plays a pivotal role in the predictive maintenance of industrial manufacturing systems. However, one major problem with existing RUL estimation algorithms is assumption single health degradation trend for different machine stages. To improve accuracy various trends, this article proposes an algorithm dubbed degradation-aware long short-term memory (LSTM) autoencoder (AE) (DELTA). First, Hilbert transform adopted to evaluate stage and factor...
In industry, prognostics and health management (PHM) is used to improve the system reliability efficiency. PHM, remaining useful life (RUL) prediction plays a key role in preventing machine failure reducing operation cost. Recently, with development of deep learning technology, long short-term memory (LSTM) convolutional neural networks (CNNs) are adopted into many RUL approaches, which shows impressive performances. However, existing learning-based methods directly utilize raw signals....
Data-driven fault diagnosis plays a key role in stability and reliability of operations modern industries. Recently, deep learning has achieved remarkable performance classification tasks. However, reality, the model can be deployed under highly varying working environments. As result, trained certain environment (i.e., distribution) fail to generalize well on data from different environments distributions). The naive approach training new for each would infeasible practice. To address this...
Learning time-series representations when only unlabeled data or few labeled samples are available can be a challenging task. Recently, contrastive self-supervised learning has shown great improvement in extracting useful from via contrasting different augmented views of data. In this work, we propose novel <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">T</b> ime- xmlns:xlink="http://www.w3.org/1999/xlink">S</b> eries representation framework...
Accurate estimation of the remaining useful life (RUL) lithium-ion batteries is critical for their large-scale deployment as energy storage devices in electric vehicles and stationary storage. A fundamental understanding factors affecting RUL crucial accelerating battery technology development. However, it very challenging to predict accurately because complex degradation mechanisms occurring within batteries, well dynamic operating conditions practical applications. Moreover, due...
Skin tissue, composed of epidermis, dermis, and subcutaneous is the largest organ human body.It serves as a protective barrier against pathogens physical trauma plays crucial role in maintaining homeostasis.Skin diseases, such psoriasis, dermatitis, vitiligo, are prevalent can seriously impact quality patient life.Exosomes lipid bilayer vesicles derived from multiple cells with conserved biomarkers important mediators intercellular communication.Exosomes skin cells, blood, stem main types...
Graph neural networks (GNN) are increasingly used to classify EEG for tasks such as emotion recognition, motor imagery and neurological diseases disorders. A wide range of methods have been proposed design GNN-based classifiers. Therefore, there is a need systematic review categorisation these approaches. We exhaustively search the published literature on this topic derive several categories comparison. These highlight similarities differences among methods. The results suggest prevalence...
Multivariate Time-Series (MTS) data is crucial in various application fields. With its sequential and multi-source (multiple sensors) properties, MTS inherently exhibits Spatial-Temporal (ST) dependencies, involving temporal correlations between timestamps spatial sensors each timestamp. To effectively leverage this information, Graph Neural Network-based methods (GNNs) have been widely adopted. However, existing approaches separately capture dependency fail to the Different sEnsors at...