layout |
---|
page |
- Computational molecular biology
We study computational problems regarding both protein-coding and non-coding genes. For coding genes, we predict subcellular localizations, type III secreted proteins and structual properties through sequence analysis and knowledge extraction from public databases.
For non-coding genes, we focus on the identification of certain types of non-coding RNAs, and differential expression analysis for microRNAs with microarray or deep sequencing data.
- Biomedical information processing
We design algorithms to screen tumor biomarkers efficiently and reliably. Now we are particularly interested in identifying microRNA biomarkers, and microRNA-disease associations.
- Machine learning and data mining
We develop algorithms to deal with large scale, imbalanced and multi-label problems, mainly from the biomedical field.