Computational Systems Biology involves studying cellular functions and its components at varying degrees of granularity. These levels range from the nano-scale molecular structures (atomic level) to entire organs such as heart and lungs (phenotype level). Our research focus is mainly on the functional aspects of cellular components, in the form of Biopathways.

Complexity Spectrum of Biological Systems

 

Biopathways

The mechanisms driving the cell are all the various chemical reactions in the cell. Aside from transcription and translation as depicted by the Central Dogma of Molecular Biology, the various organic molecules undergo other processes such as phosphorylation, dephosphorylation, translation, association and dissociation giving rise to the complex ‘circuitry’ of the cell. The possible reactions, and their reactants can be drawn in a graph-like schema, known as Biopathways.

 

It has been widely accepted that there are three main classifications of Biopathways, each depicting individual aspects of cellular function.

  • Metabolic Pathways
  • Signaling Pathways
  • Gene Regulatory Networks

Here at the Computational Systems Biology group at NUS, we are interested in modeling and analyzing Signaling Pathways and Gene Regulatory Networks.

Currently we have joint projects with the Genome Institute of Singapore and the Department of Biochemistry, NUS, modeling various pathways that are involved in important cell processes such as differentiation and apoptosis. Using these pathways as examples, we hope to be able to develop a set of tools and modeling methodology to produce accurate models that can be validated and can be used to predict new phenomena.
 

Hybrid Modeling Framework

Biology is starting to change from a qualitative, descriptive science to a quantitative, predictive one. Quantitative methods use numerical representation, such as protein concentrations and the rates of chemical reactions to describe the system. Qualitative methods, on the other hand, use non-numerical examination or interpretation of observations to discover the relationships between the components of the pathway.

Most methods of modeling biopathways are either qualitative (such as Boolean networks and Bayesian networks) or quantitative (such as Ordinary Differential Equations). Each type of modeling has its own advantages and disadvantages. Interestingly, some models of computation have both features - Hybrid Models. Currently, we are using the Hybrid Petri Net methodology to model and study biopathways. However in the future we aim to provide our own hybrid framework that will not only capture molecular interactions, but also allow different types of pathways to interact with one another.
 

Issues that are to be addressed in this research

 

Research Areas

Below is a list of biopathways and issues that we are now looking into. They are not standalone projects per se, as the techniques and framework being developed will be used to iteratively enhance the details of the pathways being modeled.

[Home]