Saving Significant Amount of Time in MD Simulations by Using an Implicit Solvent Model and Elevated Temperatures

Table 1

Lilliefors test for the different test sets. The table includes the results of the 4 sets of runs; for each set the median of the 50 runs is calculated. In addition the P values of the Lilliefors statistical test are presented for each of the 3 distributions: normal, extreme value, and exponential. The null hypothesis is that the sample comes from the given distribution. We see that the normal and extreme value distributions can be rejected (with a 5% P value), while the exponential distribution, which is the expected one, cannot.

Water modeling approach	Explicit	Explicit	Implicit	Implicit
Temperature (°K)	298	450	298	450
Median simulation dissolution time (ps)	86.4	15.6	>1000	234.853
Median real-world dissolution simulation time	57.2 min	13.4 min	>0.73 min	0.16 min
P value to reject H0: exponential	0.2808	0.1902		0.0906
P value to reject H0: normal	0.001	0.001		0.0125
P value to reject H0: extreme value	0.001	0.001		0.001