A5052900302- Computational Drug Discovery Methods
- Machine Learning in Materials Science
- Protein Structure and Dynamics
- Chemical Synthesis and Analysis
- Machine Learning in Bioinformatics
- Click Chemistry and Applications
- RNA and protein synthesis mechanisms
- vaccines and immunoinformatics approaches
- Microbial Natural Products and Biosynthesis
- DNA and Nucleic Acid Chemistry
- CRISPR and Genetic Engineering
- Genetics, Bioinformatics, and Biomedical Research
- RNA Research and Splicing
- Bioinformatics and Genomic Networks
- Tuberculosis Research and Epidemiology
- Monoclonal and Polyclonal Antibodies Research
- Animal Genetics and Reproduction
- Receptor Mechanisms and Signaling
- Biomedical Text Mining and Ontologies
- Metabolomics and Mass Spectrometry Studies
- Cell Image Analysis Techniques
- Cytokine Signaling Pathways and Interactions
- Intravenous Infusion Technology and Safety
- Biosimilars and Bioanalytical Methods
- Glycosylation and Glycoproteins Research
Central South University
2024-2025
Zhejiang University
2020-2024
Zhejiang Lab
2020-2024
Jiangxi Agricultural University
2024
Poultry Research Institute
2024
Tencent (China)
2021-2023
Zhejiang University of Science and Technology
2020-2023
Yalong Hydro (China)
2021-2022
Wuhan University
2001-2021
Wenzhou Medical University
2021
Abstract Graph neural networks (GNN) has been considered as an attractive modelling method for molecular property prediction, and numerous studies have shown that GNN could yield more promising results than traditional descriptor-based methods. In this study, based on 11 public datasets covering various endpoints, the predictive capacity computational efficiency of prediction models developed by eight machine learning (ML) algorithms, including four (SVM, XGBoost, RF DNN) graph-based (GCN,...
Accurate quantification of protein–ligand interactions remains a key challenge to structure-based drug design. However, traditional machine learning (ML)-based methods based on handcrafted descriptors, one-dimensional protein sequences, and/or two-dimensional graph representations limit their capability learn the generalized molecular in 3D space. Here, we proposed novel deep representation framework named InteractionGraphNet (IGN) from structures complexes. In IGN, two independent...
Abstract Graph neural networks (GNNs) have been widely used in molecular property prediction, but explaining their black-box predictions is still a challenge. Most existing explanation methods for GNNs chemistry focus on attributing model to individual nodes, edges or fragments that are not necessarily derived from chemically meaningful segmentation of molecules. To address this challenge, we propose method named substructure mask (SME). SME based well-established and provides an...
Abstract Although a wide variety of machine learning (ML) algorithms have been utilized to learn quantitative structure–activity relationships (QSARs), there is no agreed single best algorithm for QSAR learning. Therefore, comprehensive understanding the performance characteristics popular ML used in highly desirable. In this study, five linear [linear function Gaussian process regression (linear-GPR), support vector (linear-SVM), partial least squares (PLSR), multiple (MLR) and principal...
Safety is a main reason for drug failures, and therefore, the detection of compound toxicity potential adverse effects in early stage development highly desirable. However, accurate prediction many endpoints extremely challenging due to low accessibility sufficient reliable data, as well complicated diversified mechanisms related toxicity. In this study, we proposed novel multitask graph attention (MGA) framework learn regression classification tasks simultaneously. MGA has shown excellent...
Molecular property prediction models based on machine learning algorithms have become important tools to triage unpromising lead molecules in the early stages of drug discovery. Compared with mainstream descriptor- and graph-based methods for molecular predictions, SMILES-based can directly extract features from SMILES without human expert knowledge, but they require more powerful feature extraction a larger amount data training, which makes less popular. Here, we show great potential...
Annotating active sites in enzymes is crucial for advancing multiple fields including drug discovery, disease research, enzyme engineering, and synthetic biology. Despite the development of numerous automated annotation algorithms, a significant trade-off between speed accuracy limits their large-scale practical applications. We introduce EasIFA, an site algorithm that fuses latent representations from Protein Language Model 3D structural encoder, then aligns protein-level information with...
Artificial intelligence (AI)-aided drug design has demonstrated unprecedented effects on modern discovery, but there is still an urgent need for user-friendly interfaces that bridge the gap between these sophisticated tools and scientists, particularly those who are less computer savvy. Herein, we present DrugFlow, AI-driven one-stop platform offers a clean, convenient, cloud-based interface to streamline early discovery workflows. By seamlessly integrating range of innovative AI algorithms,...
Abstract Breast cancer resistance protein (BCRP/ABCG2), an ATP-binding cassette (ABC) efflux transporter, plays a critical role in multi-drug (MDR) to anti-cancer drugs and drug–drug interactions. The prediction of BCRP inhibition can facilitate evaluating potential drug interactions early stage discovery. Here we reported structurally diverse dataset consisting 1098 inhibitors 1701 non-inhibitors. Analysis various physicochemical properties illustrates that are more hydrophobic aromatic...
Abstract Graph neural networks (GNN) has been considered as an attractive modelling method for molecular property prediction, and numerous studies have shown that GNN could yield more promising results than traditional descriptor-based methods. In this study, based on 11 public datasets covering various endpoints, the predictive capacity computational efficiency of prediction models developed by eight machine learning (ML) algorithms, including four (SVM, XGBoost, RF DNN) graph-based (GCN,...
Abstract Accurate predictions of druggability and bioactivities compounds are desirable to reduce the high cost time drug discovery. After more than five decades continuing developments, quantitative structure–activity relationship (QSAR) methods have been established as indispensable tools that facilitate fast, reliable affordable assessments physicochemical biological properties in drug-discovery programs. Currently, there mainly two types QSAR methods, descriptor-based graph-based...
Nucleic acid (NA)-ligand interactions are of paramount importance in a variety biological processes, including cellular reproduction and protein biosynthesis, therefore, NAs have been broadly recognized as potential drug targets. Understanding NA-ligand at the atomic scale is essential for investigating molecular mechanism further assisting NA-targeted discovery. Molecular docking one predominant computational approaches predicting between small molecules. Despite availability versatile...
In this paper, we developed a novel conformation generation model, termed SDEGen, learning how molecule evolves in stochastic dynamics system starting from noise and eventually relaxing to the that falls into low energy minima.
The first study to evaluate the capability of MM/PBSA and MM/GBSA predict binding affinities recognize near-native poses for RNA-ligand systems.
Abstract The predictive performance of classical scoring functions (SFs) seems to have reached a plateau. Currently, SFs relying on sophisticated machine learning techniques shown great potential in binding affinity prediction and virtual screening. As one the most indispensable components workflow training function (MLSF), featurization or representation process enables us catch certain physical processes that are important for protein–ligand interactions obtain machine‐readable...
The molecular mechanics/generalized Born surface area (MM/GBSA) has been widely used in end-point binding free energy prediction structure-based drug design (SBDD). However, practice, it is usually being treated as a disputed method mostly because of its system dependence. Here, combining with machine-learning optimization, we developed novel version MM/GBSA, named variable atomic dielectric MM/GBSA (VAD-MM/GBSA), by assigning constants directly to the protein/ligand atoms. new strategy...
Development of accurate machine-learning-based scoring functions (MLSFs) for structure-based virtual screening against a given target requires large unbiased dataset with structurally diverse actives and decoys. However, most datasets the development MLSFs were designed traditional SFs may suffer from hidden biases data insufficiency. Hereby, we developed new approach named Topology-based Conformation-based decoys generation (TocoDecoy), which integrates two strategies to generate by...
Liver microsomal stability, a crucial aspect of metabolic significantly impacts practical drug discovery. However, current models for predicting liver stability are based on limited molecular information from single species. To address this limitation, we constructed the largest public database compounds three common species: human, rat, and mouse. Subsequently, developed series classification using both traditional descriptor-based classic graph-based machine learning (ML) algorithms....
Target identification is a critical stage in the drug discovery pipeline. Various computational methodologies have been dedicated to enhancing classification performance of compound-target interactions, yet significant room remains for improving recommendation performance. To address this challenge, we developed TarIKGC, tool target prioritization that leverages semantics enhanced knowledge graph (KG) completion. This method harnesses representation learning within heterogeneous...
Deoxyribonucleic acid (DNA) serves as a repository of genetic information in cells and is critical molecular target for various antibiotics anticancer drugs. A profound understanding small molecule interaction with DNA crucial the rational design DNA-targeted therapies. While mechanics/Poisson-Boltzmann surface area (MM/PBSA) mechanics/generalized Born (MM/GBSA) approaches have been well established predicting protein-ligand binding, their application to DNA-ligand interactions has less...
Tuberculosis (TB) is an infectious disease caused by Mycobacterium tuberculosis (Mtb) and it has been one of the top 10 causes death globally. Drug-resistant (XDR-TB), extensively resistant to commonly used first-line drugs, emerged as a major challenge TB treatment. Hence, quite necessary discover novel drug candidates for In this study, based on different types molecular representations, four machine learning (ML) algorithms, including support vector machine, random forest (RF), extreme...
Many deep learning (DL)-based molecular generative models have been proposed to design novel molecules. These may perform well on benchmarks, but they usually do not take real-world constraints into account, such as available training data set, synthetic accessibility, and scaffold diversity in drug discovery. In this study, a new algorithm, ChemistGA, was by combining the traditional heuristic algorithm with DL, which crossover of genetic (GA) redefined DL conjunction GA, an innovative...