Dongsheng Zhao

Dr. Dongsheng Zhao


Research Focus & Metrics

Academic Career Length

21Years (2006-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
18 March 2010 in BMC Bioinformatics
Verified
Journal Article
Open access
PDF file
ZSZhiyong Shen
Zhiyong Shen
WQWubin Qu
Wubin Qu
WWWen Wang
Wen Wang
YLYiming Lu
Yiming Lu
YWYonghong Wu
Yonghong Wu
CZChenggang Zhang
Chenggang Zhang
Background: Multiplex PCR, defined as the simultaneous amplification of multiple regions of a DNA template or multiple DNA templates using more than one primer set (comprising a forward primer and a reverse primer) in one tube, has been widely used in diagnostic applications of clinical and environmental microbiology studies. However, primer design for multiplex PCR is still a challenging problem and several factors need to be considered. These problems include mis-priming due to nonspecific binding to non-target DNA templates, primer dimerization, and the inability to separate and purify DNA amplicons with similar electrophoretic mobility. Results: A program named MPprimer was developed to help users for reliable multiplex PCR primer design. It employs the widely used primer design program Primer3 and the primer specificity evaluation program MFEprimer to design and evaluate the candidate primers based on genomic or transcript DNA database, followed by careful examination to avoid primer dimerization. The graph-expanding algorithm derived from the greedy algorithm was used to determine the optimal primer set combinations (PSCs) for multiplex PCR assay. In addition, MPprimer provides a virtual electrophotogram to help users choose the best PSC. The experimental validation from 2× to 5× plex PCR demonstrates the reliability of MPprimer. As another example, MPprimer is able to design the multiplex PCR primers for DMD (dystrophin gene which caused Duchenne Muscular Dystrophy), which has 79 exons, for 20×, 20×, 20×, 14×, and 5× plex PCR reactions in five tubes to detect underlying exon deletions. Conclusions: MPprimer is a valuable tool for designing specific, non-dimerizing primer set combinations with constrained amplicons size for multiplex PCR assays.

26 July 2025 in Nature Communications
Verified
Journal Article
Open access
PDF file
YZyanpeng zhao
yanpeng zhao
YXYuting Xing
Yuting Xing
YZYixin Zhang
Yixin Zhang
YWYifei Wang
Yifei Wang
MWMengxuan Wan
Mengxuan Wan
XBXiaochen Bo
Xiaochen Bo
Drug-target interaction (DTI) prediction is a crucial component of drug discovery. Recent deep learning methods show great potential in this field but also encounter substantial challenges. These include generating reliable confidence estimates for predictions, enhancing robustness when handling novel, unseen DTIs, and mitigating the tendency toward overconfident and incorrect predictions. To solve these problems, we propose EviDTI, a novel approach utilizing evidential deep learning (EDL) for uncertainty quantification in neural network-based DTI prediction. EviDTI integrates multiple data dimensions, including drug 2D topological graphs and 3D spatial structures, and target sequence features. Through EDL, EviDTI provides uncertainty estimates for its predictions. Experimental results on three benchmark datasets demonstrate the competitiveness of EviDTI against 11 baseline models. In addition, our study shows that EviDTI can calibrate prediction errors. More importantly, well-calibrated uncertainty information enhances the efficiency of drug discovery by prioritizing DTIs with higher confident predictions for experimental validation. In a case study focused on tyrosine kinase modulators, uncertainty-guided predictions identify novel potential modulators targeting tyrosine kinase FAK and FLT3. These results underscore the potential of evidential deep learning as a robust tool for uncertainty quantification in DTI prediction and its broader implications for accelerating drug discovery.

10 December 2019 in BMC Medical Informatics and Decision Making
Verified
Journal Article
Open access
PDF file
XLXuemeng Li
Xuemeng Li
DBDi Bian
Di Bian
JYJinghui Yu
Jinghui Yu
MLMei Li
Mei Li
DZDongsheng Zhao
Dongsheng Zhao
Background: With the character of high incidence, high prevalence and high mortality, stroke has brought a heavy burden to families and society in China. In 2009, the Ministry of Health of China launched the China national stroke screening and intervention program, which screens stroke and its risk factors and conducts high-risk population interventions for people aged above 40 years old all over China. In this program, stroke risk factors include hypertension, diabetes, dyslipidemia, smoking, lack of exercise, apparently overweight and family history of stroke. People with more than two risk factors or history of stroke or transient ischemic attack (TIA) are considered as high-risk. However, it is impossible for this criterion to classify stroke risk levels for people with unknown values in fields of risk factors. The missing of stroke risk levels results in reduced efficiency of stroke interventions and inaccuracies in statistical results at the national level. In this paper, we use 2017 national stroke screening data to develop stroke risk classification models based on machine learning algorithms to improve the classification efficiency. Method: Firstly, we construct training set and test sets and process the imbalance training set based on oversampling and undersampling method. Then, we develop logistic regression model, Naïve Bayesian model, Bayesian network model, decision tree model, neural network model, random forest model, bagged decision tree model, voting model and boosting model with decision trees to classify stroke risk levels. Result: The recall of the boosting model with decision trees is the highest (99.94%), and the precision of the model based on the random forest is highest (97.33%). Using the random forest model (recall: 98.44%), the recall will be increased by about 2.8% compared with the method currently used, and several thousands more people with high risk of stroke can be identified each year. Conclusion: Models developed in this paper can improve the current screening method in the way that it can avoid the impact of unknown values, and avoid unnecessary rescreening and intervention expenditures. The national stroke screening program can choose classification models according to the practice need.

6 September 2006 in BMC Bioinformatics
Verified
Journal Article
Open access
WSWenjie Shu
Wenjie Shu
XBXiaochen Bo
Xiaochen Bo
RLRujia Liu
Rujia Liu
DZDongsheng Zhao
Dongsheng Zhao
ZZZhiqiang Zheng
Zhiqiang Zheng
SWShengqi Wang
Shengqi Wang
Background: The diverse functions of ncRNAs critically depend on their structures. Mutations in ncRNAs disrupting the structures of functional sites are expected to be deleterious. RNA deleterious mutations have attracted wide attentions because some of them in cells result in serious disease, and some others in microbes influence their fitness. Results: The RDMAS web server we describe here is an online tool for evaluating structural deleteriousness of single nucleotide mutation in RNA genes. Several structure comparison methods have been integrated; sub-optimal structures predicted can be optionally involved to mitigate the uncertainty of secondary structure prediction. With a user-friendly interface, the web application is easy to use. Intuitive illustrations are provided along with the original computational results to facilitate quick analysis. Conclusion: RDMAS can be used to explore the structure alterations which cause mutations pathogenic, and to predict deleterious mutations which may help to determine the functionally critical regions. RDMAS is freely accessed via http://biosrv1.bmi.ac.cn/rdmas.

31 March 2011 in BMC Genomics
Verified
Journal Article
Open access
PDF file
WQWubin Qu
Wubin Qu
CRChanghong Ren
Changhong Ren
YLYuan Li
Yuan Li
JSJinping Shi
Jinping Shi
JZJiye Zhang
Jiye Zhang
CZChenggang Zhang
Chenggang Zhang
Background: The Ahringer C. elegans RNAi feeding library prepared by cloning genomic DNA fragments has been widely used in genome-wide analysis of gene function. However, the library has not been thoroughly validated by direct sequencing, and there are potential errors, including: 1) mis-annotation (the clone with the retired gene name should be remapped to the actual target gene); 2) nonspecific PCR amplification; 3) cross-RNAi; 4) mis-operation such as sample loading error, etc. Results: Here we performed a reliability analysis on the Ahringer C. elegans RNAi feeding library, which contains 16,256 bacterial strains, using a bioinformatics approach. Results demonstrated that most (98.3%) of the bacterial strains in the library are reliable. However, we also found that 2,851 (17.54%) bacterial strains need to be re-annotated even they are reliable. Most of these bacterial strains are the clones having the retired gene names. Besides, 28 strains are grouped into unreliable category and 226 strains are marginal because of probably expressing unrelated double-stranded RNAs (dsRNAs). The accuracy of the prediction was further confirmed by direct sequencing analysis of 496 bacterial strains. Finally, a freely accessible database named CelRNAi (http://biocompute.bmi.ac.cn/CelRNAi/) was developed as a valuable complement resource for the feeding RNAi library by providing the predicted information on all bacterial strains. Moreover, submission of the direct sequencing result or any other annotations for the bacterial strains to the database are allowed and will be integrated into the CelRNAi database to improve the accuracy of the library. In addition, we provide five candidate primer sets for each of the unreliable and marginal bacterial strains for users to construct an alternative vector for their own RNAi studies. Conclusions: Because of the potential unreliability of the Ahringer C. elegans RNAi feeding library, we strongly suggest the user examine the reliability information of the bacterial strains in the CelRNAi database before performing RNAi experiments, as well as the post-RNAi experiment analysis.

22 November 2022 in BMC Bioinformatics
Verified
Journal Article
Open access
PDF file
XZXiangwen Zheng
Xiangwen Zheng
HDHaijian Du
Haijian Du
XLXiaowei Luo
Xiaowei Luo
FTFan Tong
Fan Tong
WSWei Song
Wei Song
DZDongsheng Zhao
Dongsheng Zhao
Background: Automatic and accurate recognition of various biomedical named entities from literature is an important task of biomedical text mining, which is the foundation of extracting biomedical knowledge from unstructured texts into structured formats. Using the sequence labeling framework and deep neural networks to implement biomedical named entity recognition (BioNER) is a common method at present. However, the above method often underutilizes syntactic features such as dependencies and topology of sentences. Therefore, it is an urgent problem to be solved to integrate semantic and syntactic features into the BioNER model.Results: In this paper, we propose a novel biomedical named entity recognition model, named BioByGANS (BioBERT/SpaCy-GraphAttentionNetwork-Softmax), which uses a graph to model the dependencies and topology of a sentence and formulate the BioNER task as a node classification problem. This formulation can introduce more topological features of language and no longer be only concerned about the distance between words in the sequence. First, we use periods to segment sentences and spaces and symbols to segment words. Second, contextual features are encoded by BioBERT, and syntactic features such as part of speeches, dependencies and topology are preprocessed by SpaCy respectively. A graph attention network is then used to generate a fusing representation considering both the contextual features and syntactic features. Last, a softmax function is used to calculate the probabilities and get the results. We conduct experiments on 8 benchmark datasets, and our proposed model outperforms existing BioNER state-of-the-art methods on the BC2GM, JNLPBA, BC4CHEMD, BC5CDR-chem, BC5CDR-disease, NCBI-disease, Species-800, and LINNAEUS datasets, and achieves F1-scores of 85.15%, 78.16%, 92.97%, 94.74%, 87.74%, 91.57%, 75.01%, 90.99%, respectively.Conclusion: The experimental results on 8 biomedical benchmark datasets demonstrate the effectiveness of our model, and indicate that formulating the BioNER task into a node classification problem and combining syntactic features into the graph attention networks can significantly improve model performance.