WP3: In Silico Design and Analysis of Biological Systems


Our long-term goal is to provide biologists with an environment for tackling problems at a system level, which cannot be handled without using Information Technology. This environment will provide biologists with modelling, analysis and simulation tools capable of dealing with complex behaviours and of representing emerging properties. We thus advocate a convergence between computer science and life sciences, putting ourselves on the computational side of Systems Biology. This emerging paradigm of biology moves from the classical reductionist approach to a system level understanding of life, where unpredictable, complex behaviours show up. Essential to that shift is the ongoing change of focus in biology from structure to functionality. In our terminology, passing from structure to function amounts to equipping syntax with a behavioural semantics. In this WP we plan to develop models, languages and tools for describing, analysing and implementing in silico bio-systems, as an additional contribution of information technology to those typical research areas in current bio-informatics, such as storing, organizing and retrieving large amounts of biological data, or searching and matching DNA sequences, just to mention two well-known topics.

We will start with the development of extensions of the CLS model (see below), with particular attention to inter-process interactions and stochastic and probabilistic aspects. These primitives will reflect the way living organisms act when interacting with one another, or with the environment. Our newly tailored formalisms will share with the existing ones their formal, executable semantics, thus allowing for re-use of established theories, methods and tools. Particularly relevant are the interpreters that run the specification of a biological organism: each run can be seen as an experiment in silico. Of course, stochastic parameters are of paramount importance to that simulation, as well as any statistical tool for further analysis of the simulated behaviour.

We will design appropriate software simulators. These will help us to check the effectiveness and the correctness of our models. The parameters and behaviours of complex systems are directly affected by even little changes of the many different operating conditions. To model complex behaviours, we shall then consider models where probabilities, possibly changing over time, are associated with evolution rules. The stochastic approach we are proposing results more adequate than the one based on differential equations, when low-concentration reactants are involved in many processes active at the same time. Our ideas and proposals will be tuned, tested and validated over case studies selected in agreement with the biologists involved in the project. Experimental data will be provided by the AM fungus Gigaspora margarita, which lives in symbiosis with many plants and has the peculiar feature of hosting a population of uncultivable bacterial endosymbionts in its cytoplasm, and by the endobacterium Candidatus Glomeribacter gigasporarum that lives inside the fungus. The feasibility of our approach will be further validated by designing and implementing prototypical software tools, based on the theoretical framework developed in the first phase of the investigation.

We expect our simulation tools to contribute (and receive feedbacks) also during the process of synthesis of the final products (WP5 and WP6). Thus, we provide a support for the process of developing the new molecules and realising the green house experiments that will test the efficiency of AMs and strigolactones with respect to plant pests.

If successful, this project will confirm a general understanding in the scientific community that Information Technology will be as indispensable for biology and viceversa, as mathematics has been for physics. In any case, there will certainly be a fruitful crossfertilization between biology and computer science.