Seminar by Amitabha Bandyopadhyay
Building gene networks - a bottom up approach
Amitabha Bandyopadhyay
Department of BSBE, IIT Kanpur
Date: Friday, May 7, 2010
Time: 4:00 PM
Venue: CS102. Abstract:
When we press an electrical switch at our homes we precisely know the outcome of it i.e. what gadget will turn on or off and what will be the effect of it. However, when it comes to cells, we cannot predict the outcome, in totality, of a perturbation with any degree of accuracy. The simple reason being that unlike the circuits of our own homes we do not have a comprehensive idea of the circuitry of a cell. The fact remains that our ability to precisely predict the outcome of a perturbation of an organism or an organ or a tissue or for that matter even a single cell is rather rudimentary. One of the Holy Grails of Systems Biology is to be able to view the biological system (which may be defined at different levels) as a set of complex networks intertwined with each other and to evolve methods to estimate the outcome of any manipulation of this system.
Though the term was coined in 1993, Systems Biology as a discipline is being seriously pursued only for last 10 years. With the availability of data sets from rather large scale experiments from many laboratories across the world it is now becoming possible to get a bird's eye view of a given biological system of a chosen dimension. One such set of experimental data deals with large scale estimation of gene expression changes upon various manipulations. Most of the times the components of the system that are being manipulated are the signaling molecules or the transcription factors, the components that are relatively on top of the hierarchy. While this approach is relatively easy from an experimental perspective but the result obtained are very difficult to analyze for a variety of technical reasons.
In our laboratory we are pursuing a novel bottom-up approach in which we are starting from validated data cataloging the presence of the most downstream genes in a hierarchy. We propose to use high-end bioinformatic tools to build the network upwards from these set of genes.
I will elaborate the validity of our approach and why we believe that this approach is likely to lead us to build better gene-interaction networks. Further, I will outline experimental approaches of testing the models and fine-tuning the model.