Kinetic modelling approaches to in vivo imaging

In vivo imaging is routinely used to probe the dynamic behaviour of proteins and cellular compartments. These methods generate large, kinetically complex data sets, which often cannot be intuitively interpreted. High-density data sets contain information that can be used to probe complex systems. Cellular processes that make up complex systems can be displayed as diagrams. Translating cell-biological diagrams into kinetic models is based on standard principles of chemical kinetics. Quantitative models allow simulation of the diagram's response to a given experimental protocol and comparison of its predictions to experimental data. Comparison of model predictions and experimental data permits quantitative hypothesis testing for even very complex hypotheses. An abundance of software tools and databases is available to support this work. The numerical analysis techniques, optimization and parameter estimation, can be used to give each hypothesis its best chance to simultaneously account for all the available experimental data. Alternative hypotheses can be quickly tested. Parameter values can be extracted from experimental data to quantify the three classes of biological processes: transformation, translocation, and binding. Biophysical properties such as diffusion coefficients, rate constants, steady-state distribution ratios, fluxes and residence times are among the most useful quantities that can be obtained. Data collected using GFP-labelled proteins in living cells are particularly well suited to kinetic analysis. This is especially true in the context of the various photobleaching protocols, such as FRAP and FLIP, and will become even more profound as new developments in GFP technology become available. As cellular process diagrams become more and more complex, it becomes essential to build databases for models as well as for experimental data. The new fields of integrative bioinformatics and pathway databases lie at the intersection of kinetic modelling and database technology. The burgeoning fields of biological modelling, computational cell biology and systems biology will need a standard language for exchange of models among software tools. Nascent standards have been proposed in the form of the XML-based tools, CellML and SBML.