I am interested in the application of computational methods to understand biological systems, in particular, signaling networks in mammalian cells.
During my PhD I worked on the development of several theoretical approaches inspired in engineering concepts to study signaling networks. You can find more about that here.
Currently, I am trying to apply those methods to data coming from high-throughput proteomics to understand signaling networks (while I keep on working on the development of the methods).
As a first challenge, we found that it is not trivial to link in an efficient manner high-throughput data to mathematical models: the data has to be stored in a structure manner with additional information (metadata), processed (normalized, etc.), visualized and then exported for analysis. To facilitate these steps we have developed an open-source Matlab toolbox called DataRail.
With the data conveniently processed, sophisticated insight can be obtained with detailed, mechanistic models but, due to the large number of unknown parameter values, modeling very large networks is an arduous task. Therefore, we are using a simplified description based on Boolean logic that encapsulates the topology and causality of the network without dealing with kinetic parameters. Using data from different cell-types we are able to determine cell-specific models, which can help to identify targets for drug discovery that influence selectively cancerous cells. These methods are embedded in CellNetOptimizer (CNO), an open-source MATLAB toolbox that uses CellNetAnalyzer as simulation engine and works in concert withDataRail. We are applying this method to different data sets. As a proof of principle, application of this method to primary and transformed liver cells allowed us to uncover significant differences in the rewiring of their signaling networks.