A5102714780- Computational Drug Discovery Methods
- Bioinformatics and Genomic Networks
- Microbial Metabolic Engineering and Bioproduction
- Advanced Multi-Objective Optimization Algorithms
- Machine Learning in Bioinformatics
- Machine Learning in Materials Science
- Metaheuristic Optimization Algorithms Research
- Hydrocarbon exploration and reservoir analysis
- Personal Information Management and User Behavior
- Optimal Experimental Design Methods
- Chemical Synthesis and Analysis
- Protein Structure and Dynamics
- Crop Yield and Soil Fertility
- Dental Implant Techniques and Outcomes
- vaccines and immunoinformatics approaches
- Speech and dialogue systems
- AI in Service Interactions
- Pharmacogenetics and Drug Metabolism
- Biofuel production and bioconversion
- Ideological and Political Education
- Education Methods and Practices
- Cleft Lip and Palate Research
- Ferroptosis and cancer prognosis
- Water Treatment and Disinfection
- Image Retrieval and Classification Techniques
Wuhan University
2020-2025
Chinese University of Hong Kong, Shenzhen
2025
Tongji Hospital
2024
Huazhong University of Science and Technology
2024
Institute of Computing Technology
2024
University of Chinese Academy of Sciences
2024
Hebei GEO University
2020-2022
City University of Hong Kong, Shenzhen Research Institute
2020
Harbin Medical University
2016-2017
East China University of Science and Technology
2014
Abstract Background The progress in computer-aided drug design (CADD) approaches over the past decades accelerated early-stage pharmaceutical research. Many powerful standalone tools for CADD have been developed academia. As programs are by various research groups, a consistent user-friendly online graphical working environment, combining computational techniques such as pharmacophore mapping, similarity calculation, scoring, and target identification is needed. Results We presented...
Protein function prediction is one of the most important biological problems in field bioinformatics. The functions proteins are generally described by a series Gene Ontology (GO) terms that have hierarchical relationships. Two factors hinder effective protein using current methods: 1) they cannot well model and learn topological semantic similarity between residues GO terms, resulting huge gap; 2) predict calculating protein-level embeddings which does not effectively protein-function...
Download This Paper Open PDF in Browser Add to My Library Share: Permalink Using these links will ensure access this page indefinitely Copy URL DOI
Many efforts have been exerted toward screening potential drugs for targets, and conducting wet experiments remains a laborious time-consuming approach. Artificial intelligence methods, such as Convolutional Neural Network (CNN), are widely used to facilitate new drug discovery. Owing the structural limitations of CNN, features extracted from this method local patterns that lack global information. However, information whole sequence special domain can influence drugtarget affinity. A fusion...
Adverse drug reactions (ADRs) are responsible for failure in clinical trials and affect life quality of patients. The identification ADRs during the early phases development is an important task. Therefore, predicting potential protein targets eliciting essential understanding pathogenesis ADRs. In this study, we proposed a computational algorithm,Integrated Network Protein-ADR relations (INPADR), to infer protein-ADR based on integrated network. First, network was constructed by connecting...
Wireless charging has become increasingly popular for electric vehicles (EVs) in recent years. However, possible device damage may happen because of the sharp variations inverter output current and ZVS angle due to alignment variation between resonant coils especially while charging. To design a protection strategy, characters phase are analyzed when wireless EV leaves. The strategy is composed insufficient protection, overcurrent excessive protection. A detection circuit designed measure...
Retrosynthesis prediction is the task of deducing reactants from reaction products, which great importance for designing synthesis routes target products. The product molecules are generally represented with some descriptors such as simplified molecular input line entry specification (SMILES) or fingerprints in order to build models. However, most existing models utilize only one descriptor and simply consider a whole rather than further mining multi-scale features, cannot fully finely...
Metabolic pathways play a crucial role in understanding the biochemistry of organisms. In metabolic pathways, modules refer to clusters interconnected reactions or sub-networks representing specific functional units biological processes within overall pathway. pathway modules, compounds are major elements and various molecules that participate biochemical modules. These can include substrates, intermediates final products. Determining presence relation is essential for synthesizing new...
Selecting the available treatment for each cancer patient from genomic context is a core goal of precision medicine, but innovative approaches with mechanism interpretation and improved performance are still highly needed. Through utilizing in vitro chemotherapy response data coupled gene miRNA expression profiles, we applied network‐based approach that identified markers not as individual molecules functional groups extracted integrated transcription factor regulatory network. Based on...
Gene Ontology (GO) is a framework that utilizes series of GO terms in Directed Acyclic Graph (DAG) to describe protein functions. Proteins are typically annotated with several or dozens terms. However, existing methods often struggle simultaneously annotate multiple relevant hierarchical dependencies proteins, as they solely rely on sequences structures. To better utilize the information and improve function annotation performance, we propose Protein Structure-Label Embedding Attention...
Task-oriented dialog systems have witnessed substantial progress due to conversational pre-training techniques. Yet, two significant challenges persist. First, most primarily utilize the latest turn's state label for generator. This practice overlooks comprehensive value of labels in boosting model's understanding future generations. Second, an overreliance on generated policy often leads error accumulation, resulting suboptimal responses when adhering incorrect actions. To combat these...
Task-oriented dialog systems have witnessed substantial progress due to conversational pre-training techniques. Yet, two significant challenges persist. First, most primarily utilize the latest turn's state label for generator. This practice overlooks comprehensive value of labels in boosting model's understanding future generations. Second, an overreliance on generated policy often leads error accumulation, resulting suboptimal responses when adhering incorrect actions. To combat these...
Abstract Genome-scale metabolic models (GEMs) are powerful tools for predicting cellular and physiological states. However, there still missing reactions in GEMs due to incomplete knowledge. Recent gaps filling methods suggest directly responses without relying on phenotypic data. they do not differentiate between substrates products when constructing the prediction models, which affects predictive performance of models. In this paper, we propose a hyperedge model that distinguishes based...