Research

Due to the development of high-throughput methods and evidence-based approaches in areas like biology and medicine, substantial quantities of data are generated and collected routinely. It has been evident for some time that progress in many application areas will crucially depend on mathematical and computational techniques for data analysis, in particular automated and semi-automated techniques for data mining that can detect interesting, human-comprehensible patterns and regularities in data. Research efforts therefore focus on the development of methods for visualizing data, finding regularities, detecting clusters and unusual subgroups, discovering new classes, anomalies or outliers, and making predictions for yet unseen cases.

(Selected) Projects

Publications

Research Areas :

• Data Mining

  • Pattern Mining

    • String Mining
    • Graph Mining
    • Quantitative Association Rules 

  • Inductive Databases

• Machine Learning

  • Margin-Based Rule Learning
  • Complex Output

• Applications

  • Protein Secondary Structure Prediction Based on Frequent Patterns
    
  • Molecular Feature Mining and Structure-Activity Relationships for Non-Congeneric Compounds
  • Extending Structure-Acitvity Relationships for Non-Congeneric Compounds to Integrate Biological Information
    
  • Leveraging Chemical Background Knowledge for the Prediction of Growth Inhibition
    
  • Gene Expression Data Analysis
  • Gene Regulation Prediction

Research at the Exelixis Lab:

• Please visit the Exelixis Lab pages for more information
• Emerging Parallel Architectures
• High Performance Computing in Bioinformatics
• Molecular Evolution
• High Performance Computing for Phylogenetic Inference