Jiangwen Sun
Ph.D., University of Connecticut
M.E., Nanjing University, China
B.M., Secondary Military Medical University, China
Assistant Professor
Department of Computer Science, College of Science
Old Dominion University
Contact Information
3300 Engineering & Computational Sciences Building
Norfolk, Virginia 23529-0162
Office: E&CS 3204
Tel: (757) 683-7712
Email: jsun@odu.edu
I mainly work on machine learning approaches for the study of human disease by systemically interrogating data from multiple dimensions of life, including phenome, genome, transcriptome, epigenome, exposome, etc. More generally, I have broad interest in various machine learning and data mining techniques, particularly those that are applicable to problems in areas related to medicine and health, such as biology, drug discovery, image analysis and social networks. The overall goal of my research is to improve the precision of medicine and advance its automation.
I am an assistant professor in the Department of Computer Science within the College of Science at the Old Dominion University (ODU), where I direct the ODU Computational Systems Medicine Lab. My research is currently supported by the university and will soon be funded by federal agencies, such as National Institute of Health (NIH) and National Science Foundation (NSF).
Before Joining ODU, I was a PhD student, PostDoc and assistant research professor in the Department of Computer Science & Engineering at the University of Connecticut (UConn). During this time I worked with Dr. Jinbo Bi (Director of UConn Health Informatics Lab), Dr. Kranzler (Professor of Psychiatry, Perelman School of Medicine, University of Pennsylvania) and Dr. Xiuchun Tian (Professor of Biotechnology, UConn) on developing novel machine learning approaches to address computational problems in human medicine and related biology, including: disease subtyping considering simultaneously the phenotypic and genotypic data, outcome prediction with temporal measurements, conservative gene regulatory module identification using transcriptome profile and phenotype imputation leveraging data on genetic variants.
As a master student in the Department of Computer Science & Technology at Nanjing University, I worked on graphical models for developing novel classification methods and applied data mining techniques on database consisting of prescriptions in Chinese traditional medicine (e.g., combinations of medical use herbs).
Selected Publications (check Google Scholar for more)
-
VIGAN: Missing View Imputation with Generative Adversarial Networks
In the proceedings of the IEEE International Conference on Big Data, 2017
-
Collaborative Phenotype Inference from Comorbid Substance Use Disorders and Genotypes
In the proceedings of the IEEE International Conference on Bioinformatics and Biomedicine (BIBM), 2017
-
A Sparse Interactive Model for Matrix Completion with Side Information
In the proceedings of the Advances In Neural Information Processing Systems (NIPS), 2016
-
Multiplicative Multitask Feature Learning
In Journal of Machine Learning Research, 17(80):1-33, 2016
-
A Cross-species Bi-clustering Approach to Identifying Conserved Co-regulated Genes
Bioinformatics, 32 (12), i137-i146, 2016
-
Quantifying Feed Efficiency of Dairy Cattle for Genome-wide Association Analysis
In The proceedings of the IEEE International Conference on Bioinformatics and Biomedicine (BIBM), 2015
-
Refining Multivariate Disease Phenotypes for High Chip Heritability
BMC Medical Genomics, 8 (Suppl 3), 2015
-
An Effective Method to Identify Heritable Components from Multivariate Phenotypes
PloS One, 10 (12), 2015
-
Longitudinal LASSO: Jointly Learning Features and Temporal Contingency for Outcome Prediction
In the Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2015
-
Multi-view Sparse Co-clustering via Proximal Alternating Linearized Minimization
In the Proceedings of The 32nd International Conference on Machine Learning (ICML), 2015
-
Identifying Heritable Composite Traits from Multivariate Phenotypes and Genome-wide SNPs
In the proceedings of the IEEE International Conference on Bioinformatics and Biomedicine (BIBM), 2014
-
A Sparse Integrative Cluster analysis for Understanding Soybean Phenotypes
In the proceedings of the IEEE International Conference on Bioinformatics and Biomedicine (BIBM), 2014
-
On Multiplicative Multitask Feature Learning
In the proceedings of the Advances In Neural Information Processing Systems (NIPS), 2014
-
Transcriptional Profiles of Bovine in Vivo Pre-implantation Development
BMC Genomics, 15 (1), 2014
-
Multi-view Singular Value Decomposition for Disease Subtyping and Genetic Associations
BMC Genetics, 15 (1), 2014
-
American Journal of Medical Genetics Part B: Neuropsychiatric Genetics, 165 (2), 148-156, 2014
-
Multiview Comodeling to Improve Subtyping and Genetic Association of Complex Diseases
IEEE Journal of Biomedical and Health Informatics, 18 (2), 548-554, 2014
-
Multi-view Biclustering for Genotype-phenotype Association Studies of Complex Diseases
In the proceedings of the IEEE International Conference on Bioinformatics and Biomedicine (BIBM), 2013
-
A machine Learning Approach to College Drinking Prediction and Risk Factor Identification
ACM Transactions on Intelligent Systems and Technology (TIST), 4 (4), 2013
-
In the Proceedings of the 19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2013
-
A Multi-objective Program for Quantitative Subtyping of Clinically Relevant Phenotypes
In the proceedings of the IEEE International Conference on Bioinformatics and Biomedicine (BIBM), 2012
-
Improved Methods to Identify Stable, Highly Heritable Subtypes of Opioid Use and Related Behaviors
Addictive Behaviors, 37 (10), 2012
Multi-view Bi-clustering
This set of algorithms includes three multi-view bi-clustering methods. All three methods can be used to identify clusters from multi-view data that are with consensus from all views and simultaneously identify features from each view that are associated with the identified clusters. The implementation of all these methods are in both Matlab ard R.
Reference Papers
-
A Cross-species Bi-clustering Approach to Identifying Conserved Co-regulated Genes
Bioinformatics, 32 (12), i137-i146, 2016
-
Multi-view Sparse Co-clustering via Proximal Alternating Linearized Minimization
In the Proceedings of The 32nd International Conference on Machine Learning (ICML), 2015
-
Multi-view Singular Value Decomposition for Disease Subtyping and Genetic Associations
BMC Genetics, 15 (1), 2014
Heritable Component Analysis
This package includes two methods that identify heritable component of a complex trait (such as substance use disorder) characterized by multiple low-level phenotypes. One method is maximum likelihood based and the other one is restricted maximum likelihood based. The two methods are currently implemented with Matlab.
Reference Papers
-
An Effective Method to Identify Heritable Components from Multivariate Phenotypes
BMC Medical Genomics, 8 (Suppl 3), 2015
-
Refining Multivariate Disease Phenotypes for High Chip Heritability
PloS One, 10 (12), 2015
Program Committees
- HealthInf 2019
- HealthInf 2018
- Workshop of Machine Learning and Big Data Research for Disease Classification and Complex Phenotyping at BIBM 2017
- Workshop of Machine Learning and Big Data Research for Disease Classification and Complex Phenotyping at BIBM 2016
Ad hoc Reviewer
- Journal Entropy
- IEEE Transactions on Big Data
- Journal of Applied Mathematical Modelling
- Journal of BioMed Research International
- IEEE/ACM Transactions on Computational Biology and Bioinformatics
- IEEE Transactions on Neural Networks and Learning Systems
- Journal Neurocomputing
- Journal of Neural Computing and Applications
- Journal of Computers in Biology and Medicine
- Advances In Neural Information Processing Systems (NIPS)
Grant Reviewer
- Swiss National Science Foundation, 2017