Abstract
It has been reported recently that classical, isothermal–isobaric molecular dynamics (NTP MD) simulations at a time step of 1.00fs of the standard-mass time (Δt=1.00fssmt) and a temperature of ≤340K using uniformly reduced atomic masses by tenfold offers better configurational sampling than standard-mass NTP MD simulations at the same time step. However, it has long been reported that atomic masses can also be increased to improve configurational sampling because higher atomic masses permit the use of a longer time step. It is worth investigating whether standard-mass NTP MD simulations at Δt=2.00 or 3.16fssmt can offer better or comparable configurational sampling than low-mass NTP MD simulations at Δt=1.00fssmt. This article reports folding simulations of two β-hairpins showing that the configurational sampling efficiency of NTP MD simulations using atomic masses uniformly reduced by tenfold at Δt=1.00fssmt is statistically equivalent to and better than those using standard masses at Δt=3.16 and 2.00fssmt, respectively. The results confirm that, relative to those using standard masses at routine Δt=2.00fssmt, the low-mass NTP MD simulations at Δt=1.00fssmt are a simple and generic technique to enhance configurational sampling at temperatures of ≤340K.
Highlights
It has been reported recently that use of uniformly reduced atomic masses by tenfold can enhance configurational sampling in classical, isothermal–isobaric molecular dynamics (NTP MD) simulations at a time step (Δt) of 1.00 fs of the standard-mass time [1]
On the other hand, according to the results reported in Ref. [1], standardmass NTP MD simulations with the SHAKE algorithm could be performed at Time step (Δt) 1⁄43.16 fssmt and T r340 K without the use of the hydrogen mass repartitioning scheme that is developed to avoid system instability caused by the use of Δt Z2.00 fssmt[15,19,26]
It was reported previously that CLN025 did not fold from a fully extended backbone conformation to its native conformation in 10 unique, independent, all-atom, and classical 500-million–timestep NTP MD simulations at Δt1⁄41.00 fssmt using FF14SB at 277 K
Summary
It has been reported recently that use of uniformly reduced atomic masses by tenfold (hereafter abbreviated as low masses) can enhance configurational sampling in classical, isothermal–. [1], standardmass NTP MD simulations with the SHAKE algorithm could be performed at Δt 1⁄43.16 fssmt and T r340 K without the use of the hydrogen mass repartitioning scheme that is developed to avoid system instability caused by the use of Δt Z2.00 fssmt[15,19,26] This article reports a comparative study of the three aims using unique, independent, all-atom, and classical NTP MD simulations—each of which was performed with the SHAKE algorithm for 106 time steps—to determine relative configurational sampling efficiencies of using mass scaling factors of 1.0 and 0.1 and time steps of 1.00, 2.00, 3.16, and 3.50 fssmt. To investigate the configurational sampling efficiency in a forcefield independent manner, two forcefields were used in this study—an up-to-date general-purpose AMBER forcefield FF14SB [32] (for either explicit or implicit solvation) and a special-purpose AMBER forcefield FF12MC [33] (for explicit solvation only)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
