A5101822148- Computational Drug Discovery Methods
- Protein Structure and Dynamics
- Machine Learning in Materials Science
- Arctic and Antarctic ice dynamics
- Climate change and permafrost
- Cardiac electrophysiology and arrhythmias
- Machine Learning in Bioinformatics
- Cryospheric studies and observations
- Context-Aware Activity Recognition Systems
- Bioinformatics and Genomic Networks
- Ion channel regulation and function
- Human Pose and Action Recognition
- Receptor Mechanisms and Signaling
- Air Quality and Health Impacts
- Advanced MRI Techniques and Applications
- Gait Recognition and Analysis
- Analytical Chemistry and Chromatography
- Video Surveillance and Tracking Methods
- ECG Monitoring and Analysis
- Heme Oxygenase-1 and Carbon Monoxide
- Renin-Angiotensin System Studies
- Neural Networks and Applications
- Climate Change and Health Impacts
- Technology and Data Analysis
- Marine and coastal ecosystems
Ocean University of China
2015-2024
Beijing Microelectronics Technology Institute
2024
Qingdao National Laboratory for Marine Science and Technology
2019-2022
Institute of Oceanographic Instrumentation
2017-2022
Shandong Academy of Sciences
2017-2022
Qilu University of Technology
2022
Wenzhou Institute of Technology Testing & Calibration
2022
University of Manchester
2021
Hulunbuir University
2021
Beijing Jiaotong University
2015
Computer-aided drug design uses high-performance computers to simulate the tasks in design, which is a promising research area. Drug-target affinity (DTA) prediction most important step of computer-aided could speed up development and reduce resource consumption. With deep learning, introduction learning DTA improving accuracy have become focus research. In this paper, utilizing structural information molecules proteins, two graphs proteins are built respectively. Graph neural networks...
Accurate prediction of molecular properties is important for new compound design, which a crucial step in drug discovery. In this paper, graph data utilized property based on convolution neural networks. addition, spatial embedding layer (C-SGEL) introduced to retain the connection information molecules. And, multiple C-SGELs are stacked construct network (C-SGEN) end-to-end representation learning. order enhance robustness network, fingerprints also combined with C-SGEN build composite...
Affinity prediction between molecule and protein is an important step of virtual screening, which usually called drug-target affinity (DTA) prediction. Its accuracy directly influences the progress drug development. Sequence-based can predict according to sequence, fast be applied large datasets. However, due lack structure information, needs improved.The proposed model WGNN-DTA competent in compound-protein interaction (CPI) tasks. Various experiments are designed verify performance method...
The prediction of drug–target affinity (DTA) is a crucial step for drug screening and discovery. In this study, new graph-based model named SAG-DTA (self-attention graph affinity) was implemented. Unlike previous methods, the proposed utilized self-attention mechanisms on molecular to obtain effective representations drugs DTA prediction. Features each atom node in were weighted using an attention score before being aggregated as molecule representation. Various scoring methods compared...
The structure of a protein is great importance in determining its functionality, and this characteristic can be leveraged to train data-driven prediction models. However, the limited number available structures severely limits performance these AlphaFold2 open-source data set predicted have provided promising solution problem, are expected benefit model by increasing training samples. In work, we constructed new that acted as benchmark implemented state-of-the-art structure-based approach...
Predicting drug-target affinity (DTA) is a crucial step in the process of drug discovery. Efficient and accurate prediction DTA would greatly reduce time economic cost new development, which has encouraged emergence large number deep learning-based methods. In terms representation target proteins, current methods can be classified into 1D sequence- 2D-protein graph-based However, both two approaches focused only on inherent properties protein, but neglected broad prior knowledge regarding...
In the process of drug discovery, identifying interaction between protein and novel compound plays an important role. With development technology, deep learning methods have shown excellent performance in various situations. However, compound-protein is complicated features extracted by most models are not comprehensive, which limits to a certain extent. this paper, we proposed multiscale convolutional network that local global topological feature using different types networks. The results...
Molecular toxicity prediction is one of the key studies in drug design. In this paper, a deep learning network based on two-dimension grid molecules proposed to predict toxicity. At first, van der Waals force and hydrogen bond were calculated according different descriptors molecules, multi-channel grids generated, which could discover more detail helpful molecular information for prediction. The generated fed into convolutional neural obtain result. A Tox21 dataset was used evaluation. This...
The semi-enclosed bays impacted by heavy anthropogenic activities have weak water exchange and purification capacities. Most of the sea suffered severe eutrophication, quality deterioration, ecosystem degradation other problems. Although many countries local governments carried out corresponding environmental protection actions, evaluation their effectiveness still requires monitoring technology data support for long-term environment change. In this study, we take Yueqing Bay, fourth largest...
Exploring protein-protein interaction (PPI) is of paramount importance for elucidating the intrinsic mechanism various biological processes. Nevertheless, experimental determination PPI can be both time-consuming and expensive, motivating exploration data-driven deep learning technologies as a viable, efficient, accurate alternative. Nonetheless, most current learning-based methods regarded pair proteins to predicted possible two separate entities when extracting features, thus neglecting...
Abstract Deep learning shortens the cycle of drug discovery for its success in extracting features molecules and proteins. Generating new with deep methods could enlarge molecule space obtain specific properties. However, it is also a challenging task considering that connections between atoms are constrained by chemical rules. Aiming at generating optimizing valid molecules, this article proposed Molecular Substructure Tree Generative Model, which generated adding substructure gradually....
Accurately identifying drug-target affinity (DTA) plays a significant role in promoting drug discovery and has attracted increasing attention recent years. Exploring appropriate protein representation methods the abundance of information is critical enhancing accuracy DTA prediction. Recently, numerous deep learning-based models have been proposed to utilize sequential or structural features target proteins. However, these capture only <italic xmlns:mml="http://www.w3.org/1998/Math/MathML"...
Molecular property prediction is important to drug design. With the development of artificial intelligence, deep learning methods are effective for extracting molecular features. In this paper, we propose a multichannel substructure-graph gated recurrent unit (GRU) architecture, which novel GRU-based neural network with attention mechanisms applied substructures learn and predict properties. features extracted at node level molecule capturing fine-grained coarse-grained information....
Cardiovascular diseases are the primary cause of death humans, and among these, ventricular arrhythmias most common death. There is plausible evidence implicating inflammation in etiology fibrillation (VF). In case systemic caused by an overactive immune response, induced inflammatory cytokines directly affect function ion channels cardiomyocytes, leading to a prolonged action potential duration (APD). However, mechanistic links between cytokine-induced molecular cellular influences...
Long QT interval syndrome (LQTS) is a highly dangerous cardiac disease that can lead to sudden death; however, its underlying mechanism remains largely unknown. This study conceived investigate the impact of two general genotypes LQTS type 2, and also therapeutic effects an emerging immunology-based treatment named KCNQ1 antibody.