def conformational_space_overlap(clustering, trajectoryHandler, matrixHandler):
current = 0
traj_ranges = {}
traj_to_file = {}
total_populations = {}
for i, pdb_source in enumerate(trajectoryHandler.sources):
num_confs = pdb_source.get_info("number_of_conformations")
traj_id = "traj_%d"%i
traj_ranges[traj_id] = (current, current + num_confs -1)
traj_to_file[traj_id] = pdb_source.get_path()
total_populations[traj_id] = num_confs
current = current + num_confs
decomposed_clusters = Separator.decompose(clustering.clusters, traj_ranges)
# Get population percents for each cluster
all_traj_ids = total_populations.keys()
relative_populations = []
for cluster_id in decomposed_clusters:
dc = decomposed_clusters[cluster_id]
relative_population = []
for traj_id in all_traj_ids:
if traj_id in dc:
relative_population.append(len(dc[traj_id]) / float(total_populations[traj_id]))
else:
relative_population.append(0.)
relative_population.append(cluster_id)
relative_populations.append(tuple(relative_population))
# Sort by first traj (to 'prettify' it a bit)
relative_populations.sort()
cluster_ids = [rp[-1] for rp in relative_populations]
relative_populations = numpy.array([rp[0:len(all_traj_ids)] for rp in relative_populations])
rel_pop_per_id = {}
sm_rel_pop_per_id = {}
for i in range(len(all_traj_ids)):
# print all_traj_ids[i]
# print relative_populations.T[i]
rel_pop_per_id[all_traj_ids[i]] = list(relative_populations.T[i])
sm_rel_pop_per_id[all_traj_ids[i]] = smoothed(relative_populations.T[i])
plt.plot(relative_populations.T[i],label = traj_to_file[all_traj_ids[i]])
plt.legend()
plt.show()
# Calculate JSDs
jsds = {}
for traj_a in all_traj_ids:
jsds[traj_a] = {}
for traj_b in all_traj_ids:
jsds[traj_a][traj_b] = JSD(sm_rel_pop_per_id[traj_a],
sm_rel_pop_per_id[traj_b])
# Compile results
results = {
"id_to_path":traj_to_file,
"populations": rel_pop_per_id,
"JSD": jsds,
"cluster_ids": cluster_ids
}
return results
def test_classify(self):
decomposed = {
'0': {
'traj_A': [0, 1, 2]
},
'1': {
'traj_A': [3],
'traj_B': [8, 10]
},
'2': {
'traj_A': [4],
'traj_B': [14, 15]
},
'3': {
'traj_A': [5, 6]
},
'4': {
'traj_B': [9, 11, 12, 13, 7]
}
}
expected = {
'mixed': {
'1': {
'traj_A': [3],
'traj_B': [8, 10]
},
'2': {
'traj_A': [4],
'traj_B': [14, 15]
}
},
'pure': {
'0': {
'traj_A': [0, 1, 2]
},
'3': {
'traj_A': [5, 6]
},
'4': {
'traj_B': [9, 11, 12, 13, 7]
}
}
}
self.assertDictEqual(expected, Separator.classify(decomposed))
def test_classify(self):
decomposed = {
'0': {
'traj_A': [0, 1, 2]
},
'1': {
'traj_A': [3],
'traj_B': [8, 10]
},
'2': {
'traj_A': [4],
'traj_B': [14, 15]
},
'3': {
'traj_A': [5, 6]
},
'4': {
'traj_B': [9, 11, 12, 13, 7]
}
}
expected = {
'mixed': {
'1': {
'traj_A': [3],
'traj_B': [8, 10]
},
'2': {
'traj_A': [4],
'traj_B': [14, 15]
}
},
'pure': {
'0': {
'traj_A': [0, 1, 2]
},
'3': {
'traj_A': [5, 6]
},
'4': {
'traj_B': [9, 11, 12, 13, 7]
}
}
}
self.assertDictEqual(expected, Separator.classify(decomposed))
def conformational_space_comparison(clustering, trajectoryHandler, matrixHandler,
clustering_parameters, refinement_parameters):
# clustering = Refiner(matrixHandler,
# trajectoryHandler,
# clustering_parameters,
# refinement_parameters,
# observer).run(clustering)
traj_ranges = {}
current = 0
for i, pdb_source in enumerate(trajectoryHandler.sources):
num_confs = pdb_source.get_info("number_of_conformations")
traj_ranges["traj_%d"%i] = (current, current + num_confs -1)
current = current + num_confs
decomposed_clusters = Separator.separate(clustering.clusters, traj_ranges)
analysis = Analyzer.run(decomposed_clusters, matrixHandler.distance_matrix)
return analysis
def test_decompose(self):
traj_ranges = {"traj_A":(0,6),"traj_B":(7,15)}
clusters = [
Cluster(None,[0,1,2]),
Cluster(None,[3,8,10]),
Cluster(None,[14,4,15]),
Cluster(None,[5,6]),
Cluster(None,[7,9,11,12,13]),
]
for i in range(len(clusters)):
clusters[i].id = str(i)
decomposed = Separator.decompose(clusters, traj_ranges)
all_elements = []
for cluster_id in decomposed:
all_elements.extend(getAllElements(decomposed[cluster_id]))
expected = {
'0': {
'traj_A': [0, 1, 2]
},
'1': {
'traj_A': [3],
'traj_B': [8, 10]
},
'2': {
'traj_A': [4],
'traj_B': [14, 15]
},
'3': {
'traj_A': [5, 6]
},
'4': {
'traj_B': [9, 11, 12, 13, 7]
}
}
self.assertItemsEqual(range(16),sorted(all_elements))
self.assertDictEqual(expected, decomposed )
def test_decompose(self):
traj_ranges = {"traj_A": (0, 6), "traj_B": (7, 15)}
clusters = [
Cluster(None, [0, 1, 2]),
Cluster(None, [3, 8, 10]),
Cluster(None, [14, 4, 15]),
Cluster(None, [5, 6]),
Cluster(None, [7, 9, 11, 12, 13]),
]
for i in range(len(clusters)):
clusters[i].id = str(i)
decomposed = Separator.decompose(clusters, traj_ranges)
all_elements = []
for cluster_id in decomposed:
all_elements.extend(getAllElements(decomposed[cluster_id]))
expected = {
'0': {
'traj_A': [0, 1, 2]
},
'1': {
'traj_A': [3],
'traj_B': [8, 10]
},
'2': {
'traj_A': [4],
'traj_B': [14, 15]
},
'3': {
'traj_A': [5, 6]
},
'4': {
'traj_B': [9, 11, 12, 13, 7]
}
}
self.assertItemsEqual(range(16), sorted(all_elements))
self.assertDictEqual(expected, decomposed)
def conformational_space_comparison(clustering, trajectoryHandler,
matrixHandler, clustering_parameters,
refinement_parameters):
# clustering = Refiner(matrixHandler,
# trajectoryHandler,
# clustering_parameters,
# refinement_parameters,
# observer).run(clustering)
traj_ranges = {}
current = 0
for i, pdb_source in enumerate(trajectoryHandler.sources):
num_confs = pdb_source.get_info("number_of_conformations")
traj_ranges["traj_%d" % i] = (current, current + num_confs - 1)
current = current + num_confs
decomposed_clusters = Separator.separate(clustering.clusters, traj_ranges)
analysis = Analyzer.run(decomposed_clusters, matrixHandler.distance_matrix)
return analysis
def test_separate(self):
traj_ranges = {"traj_A": (0, 6), "traj_B": (7, 15)}
clusters = [
Cluster(None, [0, 1, 2]),
Cluster(None, [3, 8, 10]),
Cluster(None, [14, 4, 15]),
Cluster(None, [5, 6]),
Cluster(None, [7, 9, 11, 12, 13]),
]
for i in range(len(clusters)):
clusters[i].id = str(i)
expected = {
'mixed': {
'1': {
'traj_A': [3],
'traj_B': [8, 10]
},
'2': {
'traj_A': [4],
'traj_B': [14, 15]
}
},
'pure': {
'0': {
'traj_A': [0, 1, 2]
},
'3': {
'traj_A': [5, 6]
},
'4': {
'traj_B': [9, 11, 12, 13, 7]
}
}
}
self.assertDictEqual(expected,
Separator.separate(clusters, traj_ranges))
def conformational_space_comparison(clustering, matrixHandler,
trajectoryHandler, clustering_parameters,
refinement_parameters, observer):
# clustering = Refiner(matrixHandler,
# trajectoryHandler,
# clustering_parameters,
# refinement_parameters,
# observer).run(clustering)
# TODO: testing
traj_ranges = {}
current = 0
for i, pdb in enumerate(trajectoryHandler.pdbs):
traj_ranges["traj_%d" % i] = (current,
current + pdb["conformations"] - 1)
current = current + pdb["conformations"]
decomposed_clusters = Separator.separate(clustering.clusters, traj_ranges)
analysis = Analyzer.run(decomposed_clusters, matrixHandler.distance_matrix)
return analysis
def test_separate(self):
traj_ranges = {"traj_A":(0,6),"traj_B":(7,15)}
clusters = [
Cluster(None,[0,1,2]),
Cluster(None,[3,8,10]),
Cluster(None,[14,4,15]),
Cluster(None,[5,6]),
Cluster(None,[7,9,11,12,13]),
]
for i in range(len(clusters)):
clusters[i].id = str(i)
expected = {
'mixed': {
'1': {
'traj_A': [3],
'traj_B': [8, 10]
},
'2': {
'traj_A': [4],
'traj_B': [14, 15]
}
},
'pure': {
'0': {
'traj_A': [0, 1, 2]
},
'3': {
'traj_A': [5, 6]
},
'4': {
'traj_B': [9, 11, 12, 13, 7]
}
}
}
self.assertDictEqual(expected,Separator.separate(clusters, traj_ranges))
def conformational_space_overlap(clustering, trajectoryHandler, matrixHandler):
current = 0
traj_ranges = {}
traj_to_file = {}
total_populations = {}
for i, pdb_source in enumerate(trajectoryHandler.sources):
num_confs = pdb_source.get_info("number_of_conformations")
traj_id = "traj_%d" % i
traj_ranges[traj_id] = (current, current + num_confs - 1)
traj_to_file[traj_id] = pdb_source.get_path()
total_populations[traj_id] = num_confs
current = current + num_confs
decomposed_clusters = Separator.decompose(clustering.clusters, traj_ranges)
# Get population percents for each cluster
all_traj_ids = total_populations.keys()
relative_populations = []
for cluster_id in decomposed_clusters:
dc = decomposed_clusters[cluster_id]
relative_population = []
for traj_id in all_traj_ids:
if traj_id in dc:
relative_population.append(
len(dc[traj_id]) / float(total_populations[traj_id]))
else:
relative_population.append(0.)
relative_population.append(cluster_id)
relative_populations.append(tuple(relative_population))
# Sort by first traj (to 'prettify' it a bit)
relative_populations.sort()
cluster_ids = [rp[-1] for rp in relative_populations]
relative_populations = numpy.array(
[rp[0:len(all_traj_ids)] for rp in relative_populations])
rel_pop_per_id = {}
sm_rel_pop_per_id = {}
for i in range(len(all_traj_ids)):
# print all_traj_ids[i]
# print relative_populations.T[i]
rel_pop_per_id[all_traj_ids[i]] = list(relative_populations.T[i])
sm_rel_pop_per_id[all_traj_ids[i]] = smoothed(
relative_populations.T[i])
plt.plot(relative_populations.T[i],
label=traj_to_file[all_traj_ids[i]])
plt.legend()
plt.show()
# Calculate JSDs
jsds = {}
for traj_a in all_traj_ids:
jsds[traj_a] = {}
for traj_b in all_traj_ids:
jsds[traj_a][traj_b] = JSD(sm_rel_pop_per_id[traj_a],
sm_rel_pop_per_id[traj_b])
# Compile results
results = {
"id_to_path": traj_to_file,
"populations": rel_pop_per_id,
"JSD": jsds,
"cluster_ids": cluster_ids
}
return results
