# This example shows how to calculate approximate low-frequency
# modes for big proteins. For a description of the techniques,
# see
#
# K. Hinsen
# Analysis of domain motions by approximate normal mode calculations
# Proteins 33, 417 (1998)
#
# and
#
# K. Hinsen, A.J. Petrescu, S. Dellerue, M.C. Bellissent-Funel, G.R. Kneller
# Harmonicity in slow protein dynamics
# Chem. Phys. 261, 25 (2000)
#
from MMTK import *
from MMTK.Proteins import Protein
from MMTK.ForceFields import CalphaForceField
from MMTK.FourierBasis import FourierBasis, estimateCutoff
from MMTK.NormalModes import EnergeticModes
from MMTK.Visualization import view
import pylab
# Construct system
universe = InfiniteUniverse(CalphaForceField(2.5))
universe.protein = Protein('insulin.pdb', model='calpha')
# Find a reasonable basis set size and cutoff
nbasis = max(10, universe.numberOfAtoms()/5)
cutoff, nbasis = estimateCutoff(universe, nbasis)
print "Calculating %d low-frequency modes." % nbasis
if cutoff is None:
# Do full normal mode calculation
modes = EnergeticModes(universe, 300.*Units.K)
else:
# Do subspace mode calculation with Fourier basis
subspace = FourierBasis(universe, cutoff)
modes = EnergeticModes(universe, 300.*Units.K, subspace)
# Plot the atomic fluctuations in the first three non-zero modes
# for chain A
chain = universe.protein[0]
pylab.xticks(range(len(chain)),
[r.name for r in chain])
for i in range(6, 9):
f = modes[i]*modes[i]
pylab.plot([f[residue.peptide.C_alpha] for residue in chain])
# Show animation of the first non-trivial mode
view(modes[6], 15.)