80 likes | 221 Views
Monday of Week 3. Monday: force fields and Verlet integration Building a bpti simulation Wed – Friday: Making water box Running namd and analysis. General Workflow in MD. The BPTI Sims. simulation in vacuum: test simulation set up configuration files ; test stuff
E N D
Monday of Week 3 • Monday: • force fields and Verlet integration • Building a bpti simulation • Wed – Friday: • Making water box • Running namd and analysis
The BPTI Sims • simulation in vacuum: test simulation • set up configuration files ; test stuff • minimize the xray structure esp. if missing parts • solvate the system: real simulation • start with minimized structure from previous • add waters and ions • minimize • equilibration • simulate
Organization of BPTI Sims • /scratch00/BMEN662/bpti • bld: INITIAL INPUTS • topology; params; • psfgen info and scripts • xray structures; solvated systems etc.. • vac-sim: Simulation of Initial Inputs • inputs: conf, ../bld/[ psf, pdb] ../bld/parm • outpus: dcd, vel, out, coor, xsc, xst • solv-sim: Solvated simulation • inputs: confs, ../bld/[psf,pdb] ../bld/parm • outputs: dcd, vel, out, coor, xsc, xst • min-eq: these feed the next step • dyn
NAMD plot • in VMD ( a rudimentary plotting tool ) • copy simulation “.out” file to local system • start VMD • Extensions -> Namd Plot : Log File -> “.out” • Select Energy • namdenr (extract data from .out file) • plot with “your favorite plotter” • xmgrace thru xterm • excell
RMSD If the RMSD does not converge to some point then the system has not equilibrated, assuming you only calc. for protein
Maxwell-Boltzman Distrib. • load psf and VELOCITY FILE • copy the energy.tcl script from bpti/analy • vmd -dispdev none -psf bpti.solv.psf -dcd -e energy.tcl • xmgrace -nxy energy.dat • Data->Transformations -> Histograms • normalize and start,stop, #bins • Data->Transformations-> Nonlinear Curve: y = (2/ sqrt(Pi * a0 ^ $3 )) * sqrt(x) * exp (-x / a0) ao corresponds to kT in kcal/mo at 300K this is ~ 0.6 kb = 0.00198657 kcal/mol-K
Backcalculating a Force Constant • If equilibrated then equipartition holds • Each oscillator has 1/2kT of energy so for all bonds which have 1/2Ko(R-Ro)^2 then nbonds/2kT = 1/2Ko(R-R0)^2 or (R-Ro)^2*kT/nbonds = Ko of course must use the same types of bonds