Integrative Modelling

In the above sections we presented models and methodologies for the representation of static and dynamic aspects of biochemical networks. Actually, two main problems have been pointed out: (1) recent data is spread over dozens of heterogeneous, distributed, and evolving databases and data cleaning is important to ensure the consistency of data; (2) most of the work of modelling systems is still done manually, that is, single processes have to be designed and linked with each other. Nevertheless, the above sections did also show that conceptual modelling of static and dynamic views is important to understand the structure behind complex biological systems. We now want to apply computational methods to conceptual models to reconstruct models from available data sources systematically.

The main idea of integrative modelling is to include all aspects of modelling in a sequential and incremental process. Figure 11.7 shows the different levels of this cycle, whereby the modeller is controlling and responsible to decide the following questions:

• Which data sources should be used to obtain data?

• What type of objects are part of my system?

• In what type of processes these objects are involved?

• Which objects and processes are parts of a system and how does the system behave?

Fig. 11.7 The cycle of integrative modelling.

While these questions are part of the conceptual modelling, the modeller can be supported by computational methods, which are summarized under the categories data preprocessing, objects integration, process integration and systems integration.

0 0

Post a comment