Our research span four application topics using machine leaning, tensor models, kernel methods and graph Laplacian models,
- Spatial and single-cell transcriptomics: Spatial transcriptomics technologies have enabled spatially-resolved RNA profiling of single cells with cell identities and localizations for understanding cells’ organizations and functions. high-order graph-based learning, tensor factorization, and meta-analysis methods to integrate multiple knowledge graphs with single-cell and spatially resolved transcriptomic data, enabling the study of cellular organization and function in spatial contexts.
- Cancer genomics: We design graph-based learning, sequence alignment, and association rule-mining algorithms to identify biomarkers and build predictive models from diverse genomic and proteomic data.
- Phenome-genome association analysis: Development of graph-based learning algorithms for analyzing drug, disease and gene associations in the context of biological and biomedical networks.
- Protein remote homology detection: Development of string kernel algorithms and label propagation algorithms to infer the protein remote homologys and study their protein structures and functions.