Aim
Molecular dynamics are a key to understand the function of biomolecules (proteins and peptides, saccharides, lipids, DNA).
Currently datasets about molecular dynamics are huge (time series, details about spatial molecule structures) and therefore neither easy to handle nor to analyse. The project with FUB and ZIB in Berlin aims at reducing this information to a minimum required which needs efficient conceptualization and modeling techniques to pick the key parameters with which the data can be reproduced and thus abstracting from irrelevant details (e.g. behaviour of conformation changes are like bistable switches).
The goal of the project is to define a metamodel for molecular dyamics models that allow for

• efficient storage of these models (because only the parameters but not the model description will be stored) in SQL Server 2005 databases
• efficient comparisons between different molecular dynamics models
• an easier way for visualization platforms to visualize molecular dynamics models if (1) these models are standardized and (2) these models are simplified to hold only key parameters

The project promises to yield insights into how complex scientific representations can be abstracted in general. The general abstractions (here the abstractions can be applied on any kind of biomolecules in the end) will be used to implement according data representations and analysis on SQL Server 2005.
In the long term the molecular dynamic behaviour and especially those being activated and controlled by external perturbations (for example laser) can lead to developments into next generation computer hardware („Molecular electronics“).

Collaborators:

• Zuse-Institute Berlin (ZIB)
• Freie Universitaet Berlin

Back to Andre's home page