mol = sys.add_macromolecule(inseq, "VDR", initialize_apo=False)
# Import data
datasets = hxio.import_HDXWorkbench(
workbench_file, # Workbench input file
macromolecule=mol,
sequence=None, # FASTA sequence string
sigma0=sigma0
) # The initial estimate for experimental SD in % deuterium untis.
print(datasets)
# Add data to molecule states and initialize models
for s in range(len(mol.get_states())):
print(s)
state = mol.get_states()[s]
state.add_dataset(datasets[s])
output_model = state.set_output_model(
model.ResidueGridModel(state, grid_size=num_exp_bins))
sys.output.initialize_output_model_file(state, output_model.pf_grids)
sampler = sampling.MCSampler(sys, sigma_sample_level="timepoint")
# First, run a short minimization step
sampler.run(250, 0.0001)
sampler.run(annealing_steps, 1)
#sampler.residue_sampler.set_adjacency(True, 4)
sampler.run(annealing_steps, 0.5)
#sampler.residue_sampler.set_adjacency(True, 3)
sampler.run(annealing_steps, 0.3)
# This temperature tends to sit around 15% MC acceptance rate, which seems to be good.
sampler.run(nsteps, 0.3, write=True)
pep = dataset.create_peptide(inseq, 1)
for d in dat.keys():
tp = pep.add_timepoint(d)
for r in dat[d]:
tp.add_replicate(r, 1)
state.add_dataset(dataset)
sys.get_output().write_datasets()
output_model = model.ResidueGridModel(state, grid_size=num_exp_bins)
state.set_output_model(output_model)
sys.output.initialize_output_model_file(state, output_model.pf_grids)
sampler = sampling.MCSampler(sys, sigma_sample_level=None)
# First, run a short minimization step
#sampler.run(50, 0.0001)
for pep in dataset.get_peptides():
for tp in pep.get_timepoints():
#try:
i = tp.get_replicates()[0]
rep_score = -1 * math.log(
state.scoring_function.replicate_score(
tp.get_model_deuteration() / pep.num_observable_amides * 100,
tp.get_replicates()[0].deut, tp.get_sigma()))
print pep.sequence, tp.time, tp.get_model_deuteration(
) / pep.num_observable_amides * 100, tp.get_replicates(
)[0].deut, tp.get_score(), rep_score, "|", tp.get_sigma(
), -1 * math.log(
ds.calculate_intrinsic_rates()
ds.calculate_observable_rate_bounds()
ds.max_rate = ds.get_max_rate()
# Add data to molecule states and initialize models
for s in range(len(mol.get_states())):
state = mol.get_states()[s]
mod = model.ResidueGridModel(state, grid_size=num_exp_bins)
output_model = state.set_output_model(mod)
state.initialize()
sys.output.initialize_output_model_file(state, output_model.pf_grids)
sys.get_output().write_datasets()
sampler = sampling.MCSampler(sys, sigma_sample_level=None, pct_moves=100)
# Slowly cool system
print "Anneal 1"
sampler.run(annealing_steps, 100, find_temperature=False)
print "Anneal 2"
# Optional modification of the number of grid points that the residue can shift (default is 5)
# (False) allows for any bin in the sampling space to be chosen at random
for i in range(5):
sampler.run(annealing_steps, 50, find_temperature=False)
#sampler.residue_sampler.set_adjacency(True, 3)
sampler.run(annealing_steps, 20, find_temperature=False)
sampler.run(annealing_steps, 10, find_temperature=False)
sampler.residue_sampler.set_adjacency(True, 4)
sampler.run(annealing_steps, 5, find_temperature=False)
conditions=None, # A data.Conditions object specifying pH, Temp, etc... None uses a standard set of conditions
error_estimate=2.0, # The initial estimate for experimental SD in % deuterium untis.
n_fastamides=2, # Number of fast exchanging N-terminal amides
offset=offset) # Numbering offset between data and FASTA file. (positive or negative integer)
# Add data to molecule state
state.add_dataset(dataset)
state2.add_dataset(dataset)
sys.output.write_datasets()
output_model = model.ResidueGridModel(state, grid_size=num_exp_bins)
state.set_output_model(output_model)
state2.set_output_model(output_model)
sampler = sampling.MCSampler(sys, pct_moves = 20, sigma_sample_level="timepoint")
sys.output.initialize_output_model_file(state, output_model.pf_grids)
sys.output.initialize_output_model_file(state2, output_model.pf_grids)
sampler.run(nsteps, 2.0, write=True)
pof = analysis.ParseOutputFile(outputdir + "/models_scores_sigmas-Apo.dat", state)
pof2 = analysis.ParseOutputFile(outputdir + "/models_scores_sigmas-Apo2.dat", state2)
pof.generate_datasets()
pof2.generate_datasets()
#pof.calculate_random_sample_convergence()
#pof2.calculate_random_sample_convergence()
conv = analysis.Convergence(pof, pof2)
om = model.ResidueGridModel(state, grid_size=num_exp_bins)
s.set_output_model(om)
s.add_dataset(dataset)
states.append(s)
output_models.append(om)
sys.output.initialize_output_model_file(state, om.pf_grids)
#sampler = sampling.EnumerationSampler(sys)
#sampler.run(write=True)
#pof = analysis.ParseOutputFile(outputdir + "/models_scores_sigmas-Apo.dat", state)
sys.output.change_output_directory(output_dir_sample)
#sys.output.initialize_output_model_file(state, output_model.pf_grids)
sampler = sampling.MCSampler(sys)
sampler.run(10000, 2.0, write=True)
pof = analysis.ParseOutputFile(
output_dir_sample + "/models_scores_sigmas-Apo.dat", states[0])
pof1 = analysis.ParseOutputFile(
output_dir_sample + "/models_scores_sigmas-Apo1.dat", states[1])
pof2 = analysis.ParseOutputFile(
output_dir_sample + "/models_scores_sigmas-Apo2.dat", states[2])
pof3 = analysis.ParseOutputFile(
output_dir_sample + "/models_scores_sigmas-Apo3.dat", states[3])
plots.plot_residue_protection_factors([pof, pof1, pof2, pof3],
num_best_models=1000,
sort_sectors=True,
show=True)
