def create_hcstrings_states(Assignments, outfile='HCstrings_states.txt'):
SA = hct.get_StatesAssignments(Assignments)
states = SA.keys()
HCstrings_states = {}
n = 0
for state in states:
n += 1
print "Get HC strings for state %d/%d" % (n, len(states))
TrajID = SA[state].keys()
numhelix_state = []
HCstrings_states[state] = []
for trajid in TrajID:
TrajFile = '/Users/tud51931/projects/MSM/msm/ff03-hybridkcenter/sourcedata/Trajectories/trj%s_hc.lh5' % trajid
Traj = Trajectory.LoadFromLHDF(TrajFile)
HCstrings_states[state] += [
Traj['HCs'][i] for i in SA[state][trajid]
]
fn = outfile
if os.path.exists(fn):
newfn = fn + '.bck'
os.system('mv %s %s' % (fn, newfn))
print "Write HCstings of states into %s" % fn
HCfile = open(fn, 'w')
pickle.dump(HCstrings_states, HCfile)
HCfile.close()
print "Done."
def __init__(self,information,projectfile,populationfile,assignmentfile_fixed,tmatrixfile,rawdatafile):
try:
self.Info = information
self.ProjectInfo = Serializer.LoadFromHDF(projectfile)
self.Population = loadtxt(populationfile)
self.Assignments = Serializer.LoadFromHDF(assignmentfile_fixed)
self.Tmatrix = mmread(tmatrixfile)
self.StateAssignment = hct.get_StatesAssignments(self.Assignments)
self.getrawdata(rawdatafile)
except:
print "Having trouble with getting required files"
raise
def __init__(self, information, projectfile, populationfile,
assignmentfile_fixed, tmatrixfile, rawdatafile):
try:
self.Info = information
self.ProjectInfo = Serializer.LoadFromHDF(projectfile)
self.Population = loadtxt(populationfile)
self.Assignments = Serializer.LoadFromHDF(assignmentfile_fixed)
self.Tmatrix = mmread(tmatrixfile)
self.StateAssignment = hct.get_StatesAssignments(self.Assignments)
self.getrawdata(rawdatafile)
except:
print "Having trouble with getting required files"
raise
sys.path.append('/Users/tud51931/scripts/gfzhou')
import HelixCoilTools as hct
parser = argparse.ArgumentParser()
parser.add_argument('project',help="Path to ProjectInfo.h5,default=ProjectInfo.h5",default="ProjectInfo.h5")
parser.add_argument('-o','--Output',help="Output file. default=Nv.dat",default="Nv.dat")
args = parser.parse_args()
if os.path.exists('Nv.dat') :
print "Nv.dat exists!"
sys.exit()
ProjectInfo = Serializer.LoadFromHDF(args.project)
LongestTrajLength = max(ProjectInfo['TrajLengths'])
NumberOfHelix = -1*np.ones((ProjectInfo['NumTrajs'],LongestTrajLength))
print 'Calculating the Number of Helix for each trajectory......'
for i in range(ProjectInfo['NumTrajs']):
trajfile = ProjectInfo['TrajFilePath']+ProjectInfo['TrajFileBaseName']+'%d'%i+ProjectInfo['TrajFileType']
if os.path.exists(trajfile):
print '%d in %d Trajectories'%(i,ProjectInfo['NumTrajs']),trajfile
t = Trajectory.LoadFromLHDF(trajfile)
Nv = hct.compute_numhelix_trajectory(t)
NumberOfHelix[i,:len(Nv)] = Nv[:]
print "Save to %s"%args.Output
savetxt(args.Output,NumberOfHelix)
print "Done."
'/Users/tud51931/projects/MSM/msm/ff03-hybridkcenter/sourcedata/ProjectInfo.h5'
)
Counts = -1 * np.ones(
(ProjectInfo['NumTrajs'], max(ProjectInfo['TrajLengths'])))
print Counts.shape
Savepath = '/Users/tud51931/projects/MSM/msm/ff03-hybridkcenter/result/NvOfTrajectory'
plt.figure()
plt.xlabel('Steps')
plt.ylabel('Nv')
plt.hold(False)
for i in range(0, 93):
T = Trajectory.LoadFromHDF(
'/Users/tud51931/projects/MSM/msm/ff03-hybridkcenter/sourcedata/Trajectories/trj%d_hc.h5'
% i)
Hcount = hct.count_Helix(T)
plt.title('Nv-steps of Traj%d' % i)
plt.plot(range(len(Hcount)), Hcount, '.')
print 'Save figure to %s/Nvoftraj%d.png' % (Savepath, i)
plt.savefig('%s/Nvoftraj%d.png' % (Savepath, i))
Counts[i, :len(Hcount)] = Hcount[:]
Counts_ma = np.ma.array(Counts, mask=[Counts == -1])
H_mean = Counts_ma.mean(0)
H_std = Counts_ma.std(0)
print H_mean
plt.figure()
plt.plot(range(len(H_mean)), H_mean, 'b')
plt.title('AverageNv-Steps Of All Trajectories')
plt.xlabel('Steps')
import os, sys
import numpy as np
from msmbuilder import Trajectory
sys.path.append('~/scripts/gfzhou/')
import HelixCoilTools as hct
from scipy import savetxt
"""
This script is to get the number of helix from trajectories.
"""
datafile = "./numhelix_alltraj.txt"
if os.path.exists(datafile):
print "%s already exists!" % datafile
print "quit."
sys.exit()
path = "/Users/tud51931/projects/MSM/msm/ff03-hybridkcenter/sourcedata/Trajectories"
numhelix_alltraj = -1 * np.ones((100, 8000), dtype=int)
for i in range(100):
Trajfile = "%s/trj%d.lh5" % (path, i)
if os.path.exists(Trajfile):
T = Trajectory.LoadFromLHDF(Trajfile)
print "Compute number of helix for %s" % Trajfile
numhelix = hct.compute_numhelix_trajectory(T)
numhelix_alltraj[i][:len(numhelix)] = numhelix[:]
print "Save data to %s" % datafile
savetxt(datafile, numhelix_alltraj)
print "Done."
def Rgprediction():
try:
Rgs = loadtxt(
'/Users/tud51931/projects/MSM/msm/ff03-hybridkcenter/result/Rgs.dat'
)
except IOError:
print "Can't find Rgs.dat, please run CalculateRg.py first."
sys.exit()
StatesAsi = hct.get_StatesAssignments(Assignments)
Rgs_states = {}
for state in StatesAsi.keys():
for trajid in StatesAsi[state].keys():
for frame in StatesAsi[state][trajid]:
Rgs_states.setdefault(state,
[]).append(Rgs[int(trajid)][int(frame)])
states = [int(i) for i in Rgs_states.keys()]
states.sort()
mean_rg_states = []
std_rg_states = []
for state in states:
mean_rg_states.append(np.mean(Rgs_states['%d' % state]))
std_rg_states.append(np.std(Rgs_states['%d' % state]))
#savetxt('mean_numhelix_states0.dat',mean_numhelix_states)
#savetxt('std_numhelix_states0.dat',std_numhelix_states)
print mean_rg_states
P0 = np.zeros(len(Population))
for data in Assignments['Data']:
P0[data[0]] += 1
P0 = P0 / P0.sum()
populationslist = []
for k in range(140):
populationslist.append(P0)
P0 *= Tmatrix
Rgs_predicted = np.dot(np.array(populationslist),
np.array(mean_rg_states).reshape(-1, 1))
print Rgs_predicted
Rgs_predicted = Rgs_predicted.reshape(1, -1)[0]
plt.figure()
plt.plot(
np.arange(0, 7000, 50),
Rgs_predicted,
'ro',
)
plt.hold(True)
Counts_ma = np.ma.array(Rgs, mask=[Rgs == -1])
Rgs_mean = Counts_ma.mean(0)
Rgs_std = Counts_ma.std(0)
print Rgs_mean
plt.plot(range(len(Rgs_mean)), Rgs_mean, 'b')
plt.title('Rgs-steps')
plt.xlabel('Steps')
plt.ylabel('Rgs')
plt.legend(('Rgs_msm', 'Rgs_rawdata'), loc='upper left')
figname = 'Rgs_prediction_%sCluster%0.1f_tau%d.png' % (metrics, cutoff,
tau)
plt.savefig(figname)
print "Save to N%s" % figname
def GetHCStringsforTrajectory(trajectory):
if isinstance(trajectory, str):
gens = Trajectory.LoadFromLHDF(trajectory)
dihedrals = hct.ComputeDihedralsFromTrajectory(gens)
HCs = hct.ConvertDihedralsToHCStrings(dihedrals)
print HCs
def Nvprediction():
try:
mean_numhelix_states = loadtxt('mean_numhelix_states.dat')
except IOError:
StatesAsi = hct.get_StatesAssignments(Assignments)
NumHelix_states = hct.compute_numhelix_states(StatesAsi)
#print "NumHelix_states",NumHelix_states
#savetxt('NumHelix_states',NumHelix_states)
states = [int(i) for i in NumHelix_states.keys()]
states.sort()
mean_numhelix_states = []
std_numhelix_states = []
for state in states:
mean_numhelix_states.append(np.mean(NumHelix_states['%d' % state]))
std_numhelix_states.append(np.std(NumHelix_states['%d' % state]))
savetxt('mean_numhelix_states.dat', mean_numhelix_states)
savetxt('std_numhelix_states.dat', std_numhelix_states)
#plt.figure()
#plt.errorbar(states,mean_numhelix_states,std_numhelix_states)
#plt.xlabel("State ID")
#plt.ylabel("Number of Helix")
#plt.savefig("Numhelix_states")
#plt.show()
P0 = np.zeros(len(Population))
for data in Assignments['Data']:
P0[data[0]] += 1
P0 = P0 / P0.sum()
populationslist = []
for k in range(140): # tau = 50, so 140*50 = 7000
populationslist.append(P0)
P0 *= Tmatrix
numhelix = np.dot(np.array(populationslist),
np.array(mean_numhelix_states).reshape(-1, 1))
print numhelix
numhelix = numhelix.reshape(1, -1)[0]
plt.figure()
plt.plot(
np.arange(0, 7000, 50),
numhelix,
'ro',
) # tau = 50, so 140*50 = 7000
plt.hold(True)
Counts = -1 * np.ones(
(ProjectInfo['NumTrajs'], max(ProjectInfo['TrajLengths'])))
print Counts.shape
for i in range(0, 93):
T = Trajectory.LoadFromHDF(
'/Users/tud51931/projects/MSM/msm/ff03-hybridkcenter/sourcedata/Trajectories/trj%d_hc.h5'
% i)
Hcount = hct.count_Helix(T)
Counts[i, :len(Hcount)] = Hcount[:]
Counts_ma = np.ma.array(Counts, mask=[Counts == -1])
H_mean = Counts_ma.mean(0)
H_std = Counts_ma.std(0)
print H_mean
plt.plot(range(len(H_mean)), H_mean, 'b')
plt.title('Nv-steps')
plt.xlabel('Steps')
plt.ylabel('Nv')
plt.legend(('Nv_msm', 'Nv_rawdata'), loc='upper left')
figname = 'Nv_prediction_%sCluster%0.1f_tau%d.png' % (metrics, cutoff, tau)
plt.savefig(figname)
print "Save to %s" % figname
import os, sys
sys.path.append('~/scripts/gfzhou/')
import HelixCoilTools as hct
from msmbuilder import Trajectory
"""
This script shows how to create new trj files with hc strings.
"""
path = "/Users/tud51931/projects/MSM/msm/ff03-hybridkcenter/sourcedata/Trajectories"
for i in range(0, 100):
Trajfile = "%s/trj%d.lh5" % (path, i)
if os.path.exists(Trajfile):
T = Trajectory.LoadFromLHDF(Trajfile)
hct.CreateTrajFileWithHCstrings(T)
print "Done."
from msmbuilder import Serializer
cutoff = 3.0
metrics = 'rmsd'
if metrics.lower() == 'dihedral':
Path = "/Users/tud51931/projects/MSM/msm/ff03-dihedralhybrid/"
metrics = 'Dihedral'
elif metrics.lower() == 'rmsd':
Path = "/Users/tud51931/projects/MSM/msm/ff03-hybridkcenter/"
metrics = 'RMSD'
Path = os.path.join(Path, '%sCluster%0.1f' % (metrics, cutoff))
AssignmentFile = os.path.join(Path, "Data", "Assignments.h5")
A = Serializer.LoadFromHDF(AssignmentFile)
StateAssignment = hct.get_StatesAssignments(AssignmentFiles=A)
RMSDFile = os.path.join(Path, "Data", "RMSD.h5")
RMSD = Serializer.LoadFromHDF(RMSDFile)
rmsd_allstates = {}
for state in StateAssignment.keys():
rmsd_singlestate = []
for trajid in StateAssignment[state].keys():
rmsd_singlestate += list(
RMSD['Data'][int(trajid)][StateAssignment[state][trajid]])
rmsd_allstates[int(state)] = rmsd_singlestate
maxstatelength = max([len(i) for i in rmsd_allstates.values()])
StateRMSDs = copy.deepcopy(RMSD)
StateRMSDs['Data'] = -1 * np.ones((len(rmsd_allstates), maxstatelength))
for state in rmsd_allstates.keys():
statelength = len(rmsd_allstates[state])