Novo Nordisk supports research in mathematical biology

The Novo Nordisk Foundation is funding projects in bioscience and basic biomedicine “to promote Danish fundamental research at high international level”. The foundation supports projects relevant to understanding the human organism and/or basic mechanisms underlying health and disease.

At the Department of Mathematical Sciences, the research group Mathematical Biology last year obtained a grant for the project "Understanding bistability underlying acute and chronic infections in Pseudomonas Aeruginosa". This year professor Carsten Wiuf receives a 1.3 million DKK research grant for the project “Model Reduction in Systems Biology“.

The grant makes it possible for Carsten Wiuf to employ a postdoc for two years, starting August 2020.

Automated model reduction

Carsten Wiuf describes the project as follows:

Models of realistic cellular systems often have many variables. In many cases, it is only possible to measure experimentally a few of these. This is, for example, the case for signalling networks that typically contain many intermediate reactions and intermediate molecules before a signal is released to a response regulator. The intermediate reactions are typically fast (compared to other reactions), their molecular components short-lived and transient, and hard to measure experimentally. To build a mathematical model of such a system, it is therefore desirable to build a model without the intermediate parts.

It is the hypothesis of this proposal that the natural way of doing so is to start from a model of the full system and argue by reduction, what the simplified model should be.

The proposal sets out to find an automated method, based on graphical conditions to reduce a deterministic ODE system of a reaction network to a simpler ODE system of smaller reaction network, without the intermediate species, and to assess the validity (in dynamical terms) of the reduction at the same time.

This approach has several advantages. First of all, being automated, it is easy for the researcher to apply, even if (s)he does not have mathematical experience. Secondly, it avoids having to check mathematical conditions manually. Thirdly, in many examples in the literature, a reduction is done without addressing the validity of it. That is, to check whether the ODE system of the reduced network is a good approximation (in some sense) of the original ODE system. The proposed method will do this check automatically.

As a second aim, the proposal will make initial steps towards understanding similar reduction techniques for stochastic systems of reaction networks and the feasibility of the techniques.

Topics