Modeling techniques, implementation/selection

Besides the mathematical improvements, the atmospheric model has been adapted to a semi-Lagrangian formulation. By following selected air masses, we avoid commitments of large quantities of memory and the incursions of artificial diffusion errors. Most important, we do not end up with stacks of computer printout that relate to regions where there are no measurements. Also, predictive calculations will become more useful, but our present levels of resources and sophistication demand that effort be concentrated on verification. Only in this way can the confidence be built that is needed for applying modeling techniques to implementation planning. 

The LIN method (described below) was constructed on the premise of filtering out the high-frequency motion by NM analysis and using a large-timestep implicit method to resolve the remaining motion components. This technique turned out to work when properly implemented for up to moderate timesteps (e.g., 15 Is) [73] (each timestep interval is associated with a new linearization model). However, the CPU gain for biomolecules is modest even when substantial work is expanded on sparse matrix techniques, adaptive timestep selection, and fast minimization [73]. Still, LIN can be considered a true long-timestep method. 

Frequency-selective REDOR (fsREDOR) is a very powerful technique developed for the study of 13C and 15N uniformly labeled peptides or proteins [92]. The basic idea of this technique is to combine REDOR and soft n pulses to recouple a selected 13C-15N dipole-dipole interaction in a multiple-spin system. Usually one could use Gaussian shaped pulses to achieve the required selective n inversions. Other band selective shaped pulses have been developed for a more uniform excitation profile [93]. In its original implementation, fsREDOR was used to extract the intemuclear distances of several model crystalline compounds [92], In the past few years, this technique has proven to be very useful for the study of amyloid fibrils as well. For the Ure2p10 39 fibril samples containing 13C and 15N uniformly... 

One very important property in solvent selection is the activity coefficient. Many techniqnes exist for estimating activity coefficients (Fredenslund et al., 1977). In addition to these detailed techniques, a number of simpler approaches have been found to be very effective. These include infinite dilntion activity coefficients (Thomas and Eckert, 1984), critical solution temperatures (Francis, 1944), and solubility parameters (Barton, 1983). In implementing the above system the authors chose to use a three term solubility parameter model. 

Solving the nonlinear mathematic models by a numerical method depends on the trial values that are chosen by experience and is not based upon a particular theory. The results obtained are greatly influenced by the initial values selected that is a characteristic of strong nonlinear problems. By using numerical techniques to solve nonlinear models, iterations must be implemented. With the incorrect selection of a trial value, divergence in solution can appear. The accuracy of the numerical results cannot be estimated theoretically for nonlinear problem since there is no analytical solution available. This is why an approximate analytical solution is extremely useful for a theoretical analysis of nonlinear problem If an approximate analytical solution can be obtained, then this has a number of benefits ... 

The techniques that are to be implemented will determine what equipment is necessary. Some equipment may already be available in or near the laboratory, but it is likely that other items will have to be purchased. Once it is clear that a certain instrument must be purchased, there are still many factors that will affect selection of the exact make and model best suited to a particular laboratory. Some key issues to keep in mind are the following. Is the model in question adequate to support the contemplated technique ) Will other techniques require the same or similar instruments Will the particular model support those techniques as well Is a more versatile model available Does its versatility justify any additional cost Can the instrument (and its associated costs) be shared How reliable is the model in question Are support and maintenance available from the vendor or other sources How expensive is maintenance Will changing technology or new research goals make a different model significantly more desirable in the near future ... 

---