Run multiple simulations

Optimization running multiple simulations and testing all combinations of strategies to find the optimal solution. 

Though the theory is old, it has received little use for molecular systems due to the computational demands of running multiple simulations for significant perturbations. However, we were stimulated by the paper of Tembe and McCammon to test the method on a substantial mutation in aqueous solution. Consequently ethane and methanol were interconverted in water, and the difference in free energies of hydration was computed. The precision... 

Run multiple simulations at different reduced densities (start at p =0.1 and increase to p = 0.8). Then do the following ... 

Another approach is to run multiple MD simulations with different initial conditions. The recent work by Karplus and coworkers [19] observes that multiple trajectories with different initial conditions improve conformational sampling of proteins. They suggest that multiple trajectories should be used rather than a single long trajectory. 

In general, will scale as A p and thus can only be eliminated by increasing the number of notional particles. The quantity ( < > X NptM) can be estimated by running multiple independent simulations with fixed Np and M (Xu and Pope 1999). 

To see if monitoring these variables acmally yields any useful information on the folding pathway, multiple simulated annealing runs with all parameter monitoring facihties implemented were carried out on a family of vitamin D molecules (Figure 4) [23]. The classical biological responses for l,25(OH)2-vitamin D3 are now well established. The vitamin D hormone is known to mediate intestinal calcium absorption and bone calcium metabolism via its receptor. More recently, non-traditional activities have been discovered including cell differentiation and prohferation. 

When a model is used for performing multiple simulations, the management of simulations becomes a major issue. At a minimum, the simulations should be set up such that the outputs from different model runs can easily be identihed. The simulation setup should allow a subset of model runs to be repeated without needing to repeat all the model runs. This is often achieved by providing a separate, appropriately named folder for storing inputs, outputs, and conhguration hies for each model run, and aggregating the model runs at the end of the simulation. Such an approach allows the model runs to be performed on a distributed cluster of machines. 

Multiple simulation runs are also rrsed for parametric studies, in which one design variable at a time is systematically changed and the performance of the system is ejqrressed as a function of that parameter. Sets of sensitivity cmves are thrrs produced, providing useful advice to the designer. 

Design of multiple gates with timed injectors. To reduce the injection pressure and/or mold filling time, the gates may be opened/closed sequentially. This approach can be explored manually with multiple simulation runs or coupled with optimization routines such as neural networks or genetic algorithms. - ... 

Run multiple steady-state simulations across wide ranges of controller set points to assess the profitability of the candidate base level designs. 

MD simulations involve numerical determinations of individual trajectories, that is, solutions to Newton s equation of motion for particular initial conditions. In chemistry, this technique began with the study of individual gas-phase collisions, where large numbers of trajectories were run to explicitly average over initial conditions. For macromolecules or liquid simulations, however, the frequency of atomic collisions becomes so great that simulations often appear to be ergodic, such that a single trajectory samples phase space with the same distribution as do multiple simulations with randomized starting points. This implies that a dynamics simulation can be used to explore phase space and make connections to classical thermodynamics and kinetics. 

---