Comparative Study of NAMD and GROMACS Yanbin Wu
Comparative Study of NAMD and GROMACS Yanbin Wu, Joonho Lee and Yi Wang Team Project for Phy 466 May 11, 2007
Outline o Motivation o Simulation Set-up o Procedure o Result and analysis o Conclusion
Motivation (1) o NAMD and GROMACS: developed in the Netherlands. Fast, free. NAMD: developed in Urbana, IL. Parallel, fast for big systems, free.
Motivation (2) o Compare two packages n n n Both are widely used MD packages. Different code implementation in the two packages may cause different results Generally, one group mainly uses one package Good chance to compare two packages !!!
Simulation Setup (1) running parameter Implementation Simulation Package model (topological) force field RESULTS
Simulation Setup (2) o Algorithm n o Initial system n n o LJ parameter Model n o Size, composition Coordinate, velocity Force field n o Running parameter Charge, bonding, angle parameter Code implementation n Black box
Procedure o Find a zero point n n o Compare different water models n n o NVE NVT TIP 3 P SPCE Compare different temperature control schemes n n n Langevin Nose Hoover Berendsen
Simulation system n The simplest system o Water n n n o n Solvent of life Simple & isotropic Rich in experimental data Water+Ions Ensembles o o NVE NVT n Langevin, Nose-hoover, Berendsen
Water Model: SPCE SPC/E rigid model (Berendsen et al. , 1987) o q(h) = 0. 4238, q(O) = - 0. 8476 o O-H distance = 1 (Å) o H-O-H angle = 109. 47 ° o LJ parameter n A = 0. 37122(k. J/mol)1/6. nm and B = 0. 3428 (k. J/mol)1/12. nm
Water Model: TIP 3 P flexible model (Mahoney and Jorgensen, 2000) o q(h) = 0. 417, q(O) = - 0. 834 o O-H distance = 0. 9572 (Å) o H-O-H angle = 104. 52 ° o LJ parameter n = 0. 1521, =3. 15061(Å)
Temperature control schemes n Langevin o n Nose-hoover o n Introduce a random force and friction coefficient Introduce a thermal reservoir and a friction term in the eq. of motion Berendsen o Weak coupling first-order kinetics to an external heath bath with a given temperature
Results and Analysis (1) o Zero point n 1) 2) n NVE 0. 25 ns NVT to bring temperature up to 300 K. 1 ns NVE. Important: start with the same velocity and coordinate in both packages. NVT 1 ns NVT using Langevin dynamics temperature control, damping coefficient 5/ps.
NVE Zero Point GROMACS NAMD Water model Tip 3 p Diffusion coefficient (10 -5 cm 2/s) 4. 9799 (+/- 0. 0012) 5. 278 (+/-0. 012) 5. 6% difference
NVT Zero Point GROMACS NAMD Water model Tip 3 p Temperature control scheme Langevin (5) Diffusion coefficient (10 -5 cm 2/s) 2. 5676 (+/-0. 00009) 2. 769 (+/-0. 002) 7. 3% difference
Results and Analysis (2) o Different water models n n n 1 ns NVT using Langevin dynamics with = 5/ps. Two different water models: SPCE and TIP 3 P. Most commonly used water models. SPCE TIP 3 P 5 2. 5758 (+/- 0. 00013) 2. 5676 (+/-0. 00009) 0. 032% difference *SPCE and TIP 3 P water models exhibit similar dynamic properties in GROMACS.
Results and Analysis (3) o Damping in Langevin Dynamics n 1 ns NVT via Langevin dynamics with = 1, 5, 10 /ps. GROMACS (TIP 3 P) NAMD (TIP 3 P) 1 3. 7263 (+/-0. 00063) 4. 259 (+/-0. 025) 5 2. 5676 (+/-0. 00009) 2. 769 (+/-0. 002) 10 1. 8515 (+/- 0. 00012) 1. 958 (+/-0. 003) *Damping affects the diffusion of water dramatically. * =5/ps best reproduces the experimental result.
Results and Analysis (4) o Different temperature control schemes n n n 1 ns NVT using Langevin dynamics with = 5/ps 1 ns NVT using Nose-hoover thermostat with =0. 1 ps 1 ns NVT using Berendsen thermostat with =0. 1 ps GROMACS(SPCE) Diffusion coefficient (10 -5 cm 2/s) Langevin 2. 5676 (+/- 0. 00009) Berendsen 2. 6539 (+/- 0. 00035) Nose-Hoover 2. 5584 (+/- 0. 00016) *Different temperature control schemes can achieve similar results with well-chosen parameters.
Results and Analysis (5) n Water in the water+ion system Package Temperature control scheme water model / force field Diffusion coefficient (10 -5 (cm 2/s)) NAMD Langevin (5) tip 3 p (namd) / CHARMM 2. 2247 (+/- 0. 0015) Berendsen spce(gromacs)/ GROMACS 2. 3026 (+/- 0. 00027) Berendsen spce(reference) / reference 2. 4140 (+/- 0. 00006) Nose-hoover spce(gromacs)/ GROMACS 2. 3235 (+/- 0. 00024) Nose-hoover spce(reference) / reference 2. 4362 (+/- 0. 00036) GROMACS
Results and Analysis (5) n Na+ in the water+ion system Package Temperature control scheme water model / force field Diffusion coefficient (10 -5 (cm 2/s)) NAMD Langevin (5) tip 3 p (namd) / CHARMM 0. 8200 (+/- 0. 0250) Berendsen spce(gromacs)/ GROMACS 0. 9194 (+/- 0. 00120) Berendsen spce(reference) / reference 1. 3675 (+/- 0. 00069) Nose-hoover spce(gromacs)/ GROMACS 1. 0353 (+/- 0. 00021) Nose-hoover spce(reference) / reference 1. 2700 (+/- 0. 00033) GROMACS
Results and Analysis (5) n Cl- in the water+ion system Package Temperature control scheme water model / force field Diffusion coefficient (10 -5 (cm 2/s)) NAMD Langevin (5) tip 3 p (namd) / CHARMM 1. 1125 (+/- 0. 0130) Berendsen spce(gromacs)/ GROMACS 1. 3300 (+/- 0. 00064) Berendsen spce(reference) / reference 1. 6091 (+/- 0. 00100) Nose-hoover spce(gromacs)/ GROMACS 1. 3517 (+/- 0. 00050) Nose-hoover spce(reference) / reference 1. 4956 (+/- 0. 00061) GROMACS
Results and Analysis (5) n Radial distribution of oxygen - oxygen
Results and Analysis (5) n Radial distribution of oxygen – Cl-
Results and Analysis (5) n Radial distribution of oxygen – Na+
Conclusion o o The two packages GROMACS and NAMD produce similar results (within tolerance) using the same set of parameters. Damping coefficient affects the dynamics significantly and has to be chosen with caution. Different temperature control schemes may generate similar dynamic properties. Two different water models, SPCE and TIP 3 P were compared and only minor difference was observed regarding the diffusion of water.
Discussion o Energy conservation in NVE simulations n n n o Damping coefficient in NVT simulations n o Neighbor list update frequency Switch or shift function is required, instead of cutoff PME Balance temperature fluctuation and disturbance to the motion of the system. Different temperature control schemes
Zero point (water+ion, NVT) GROMACS NAMD Forcefield CHARMM watermodel(bonding) tip 3 p(namd) Temperature control scheme Langevin (5) Water 1. 6556 (+/-0. 00020) 2. 2247 (+/-0. 0006) Na+ 0. 5409 (+/-0. 00031) 0. 8200 (+/-0. 0250) Cl- 0. 9766 (+/-0. 00034) 1. 1125 (+/-0. 0130) Diffusion coefficient 10 -5 (cm 2/s)
Water Model (3) Package NAMD GROMACS water model TIP 3 P (NAMD) 2. 4895 e-3 2. 4248 e-6 O/H 3. 1540 e-9 1. 2921 e-17 TIP 3 P (GROMACS) 2. 4889 e-2 2. 4352 e-6 SPCE (GROMACS) 2. 6171 e-3 2. 6331 e-6 SPCE (reference) 2. 6341 e-3 2. 6679 e-6
NVE Zero Point GROMACS NAMD Forcefield CHARMM watermodel(bonding) tip 3 p(namd) Diffusion coefficient 10 -5 (cm 2/s) 2. 4890 (+/-0. 00096) 5. 278 (+/-0. 012)
- Slides: 28