Computational and systems approaches to drug discovery and development


A significant proportion of late phase attrition of new drug candidates is considered to be due to inappropriate target selection. Computational methods can potentially improve target identification and validation at the earliest possible stage of drug development by providing a better understanding of the biology of the target network. Such a knowledge should also lead to new therapeutic strategies based on molecular mechanisms.


We carry out bioinformatics and computational biology research to assist drug discovery and development. Our work aims to provide a systems view of how genes and proteins interact with each other and perform their functions in a specific biological pathway, and of how they can be perturbed by therapeutic agents.

Research themes

1) Methods development for improved efficacy and safety:
◾Integrated database for genes, proteins and chemical compounds.
◾Network-based prioritization of candidate genes/proteins.
◾Resources for biomarker discovery and toxicity prediction.
◾Prediction of protein structure, function and interaction.

2) Applications to specific systems:
Analysis of experimental data in areas including infectious diseases, cancer, vaccine development, neurological disorders, chronic inflammatory diseases and membrane proteins.


In specific drug discovery projects, our predictions have been verified experimentally and led to new therapeutic strategies. Some of our databases and software tools have been commercialized and used widely to characterize drug targets and related proteins.