Canal-U

Reasoning over large-scale biological systems with heterogeneous and incomplete data

Data produced by the domain of life sciences in the next decade are expected to be highly challenging. In addition to scalability issues which are shared with other applications domains, data produced in life sciences have very specific characteristics: multi-scale, incomplete, heterogeneous but somehow interdependent. This makes data-mining methods less efficient than expected to assist knowledge discovery. An example of such limitations is the study of biological systems in molecular and cellular biology, which cannot be uniquely identified with the data at hand. In this talk, we will introduce a strategy to study biological systems in the framework of incomplete data.
This strategy relies on reasoning and logical programming technics, allowing to model interactions within a system, take into account information carried by the overapproximated dynamics of the system, and finally extract relevant properties by solving combinatorial problems. We will illustrate this approach on the emerging field of systems ecology which aims at understanding interactions between a consortium of microbes and a host organism.