Method for inferring kinetics and network architecture (MIKANA)


My primary research is developing suitable computational methodology to extract mechanistic information of biochemical pathways from data obtained from high throughput experiments. Understanding biochemical pathways are fundamental in addressing key problems in biomedical engineering. This has the potential to create valuable treatment strategies and in the development of new drugs. High-throughput experiments produce data which are fundamental for the mapping of biochemical pathways. Nuclear magnetic resonance (NMR), mass spectrometry (MS), fluorescence spectroscopy / microscopy, and fluorescence labeling combined with autoradiography on 2-D gels are some of the techniques which allows simultaneous measurement of several metabolites. With these techniques, we can obtain measures of metabolite concentrations at any instant or progressively over time. The later is called as time series data.

Time series data is unique because it has dynamic information about the system. Several computational methods are being developed to infer reaction mechanisms from time series data . Since more data are produced new algorithmic strategies are serious requirements in order to extract information. A combination of fundamental biochemical theory and a suitable computational method will aid the determination of mechanisms from time series data. I am precisely interested in developing such a methodology. With this as my research objective, I have developed a method to infer kinetics and network architecture (MIKANA).

The method MIKANA is based on global non-linear modeling, which identifies elementary chemical reaction steps that constitute the biochemical pathway. Elementary reaction steps are those that cannot be decomposed to reveal reaction intermediates that might themselves be identified as separate chemical entities on a biochemically relevant timescale. MIKANA depends on the selection of appropriate chemical reactions from a dictionary of elementary chemical reactions, in order to best represent experimental data. This process is aided by a cost function that penalizes the use of too many reactions to fit the data. This function is named Information Criterion (IC).

My research has been very successful in determining several simple chemical reaction mechanisms, enzyme reaction mechanisms and the complete structure of a glycolytic pathway of a bacterium. Further, I have applied this method to optimize certain experimental design factors that can be implemented by the experimentalists. This work has explicitly revealed certain modification to be done in experiments in order to obtain accurate information about the system. This is a very important result to the experimentalists.

Collabotrators: Dr. Edmund J. Crampin (Auckland Bioengineering Institute) , Dr. Patrick McSharry (University of Oxford) and Márcio Mourão (PhD student).