def test_bad_tree(self):
bad_trees = []
tree_file = "tests/data/species_trees/101g_nucl_conc_unconst.combined.nwk.tre"
with open(tree_file) as f:
bad_trees.append(f.read())
good_nexus = "#NEXUS\n\n"
good_nexus += "Begin taxa;\n"
good_nexus += " Dimensions ntax=3;\n"
good_nexus += " Taxlabels\n"
good_nexus += " spc01\n"
good_nexus += " spc02\n"
good_nexus += " spc03\n"
good_nexus += " ;\n"
good_nexus += "End;\n"
good_nexus += "Begin trees;\n"
good_nexus += " Translate\n"
good_nexus += " 1 spc01,\n"
good_nexus += " 2 spc02,\n"
good_nexus += " 3 spc03\n"
good_nexus += " ;\n"
good_nwk = "tree TREE1 = ((1[&dmv={0.1}]:1,2[&dmv={0.2}]:1)[&dmv={0.3}]"
good_nexus += "End;\n"
bad_trees.append(good_nexus.replace("#NEXUS", "#NEXU"))
bad_trees.append(good_nexus.replace("#NEXUS", "NEXUS"))
bad_trees.append(good_nexus.replace("tree TREE1", "tre TREE1"))
bad_trees.append(good_nexus.replace("End;", ""))
bad_trees.append(good_nexus.replace("Translate", "T"))
bad_trees.append(good_nexus.replace("2 spc02,", "2 spc02"))
bad_trees.append(good_nexus.replace("2 spc02,", "2 spc02 asdf,"))
bad_trees.append(good_nexus.replace("2 spc02,", "2 spc03,"))
bad_trees.append(good_nexus.replace("2 spc02,", "spc02 2,"))
bad_trees.append(good_nexus.replace("2 spc02,", "asdf2 spc02,"))
bad_trees.append(good_nexus.replace("2 spc02,", "spc02; 2,"))
bad_trees.append(good_nexus.replace(";\n", ""))
bad_trees.append(good_nexus.replace("Taxlabels", "Begin trees;"))
bad_trees.append(good_nexus.replace("dmv", "emv"))
bad_trees.append(good_nexus.replace("[", ""))
bad_trees.append(good_nexus.replace("[", "").replace("]", ""))
bad_trees.append(good_nexus.replace("=", ""))
bad_trees.append(good_nexus.replace("Begin taxa", "Begin trees"))
bad_trees.append(good_nexus.replace("Begin trees", "Begin taxa"))
bad_trees.append(good_nexus.replace("[&dmv={0.5}]", ""))
bad_trees.append(good_nexus.replace("[&dmv={0.1}]", ""))
bad_trees.append(good_nexus.replace("[&dmv={0.1}]", "[&dmv={asdf}]"))
bad_trees.append(good_nexus.replace(":1,2[&dmv", ":1, 2[&dmv"))
bad_trees.append(good_nexus.replace(good_nwk, good_nwk + good_nwk))
good_generation_time = 5
for bad_tree in bad_trees:
with self.assertRaises(ValueError):
msprime.parse_starbeast(tree=bad_tree,
generation_time=good_generation_time)
def test_12_species(self):
with open("tests/data/species_trees/91genes_species_rev.tre") as f:
good_tree = f.read()
good_branch_length_units = "myr"
good_generation_time = 5
parsed_tuple = msprime.parse_starbeast(
tree=good_tree,
branch_length_units=good_branch_length_units,
generation_time=good_generation_time,
)
self.assertEqual(len(parsed_tuple), 2)
self.assertIs(type(parsed_tuple[0]), list)
self.assertEqual(len(parsed_tuple[0]), 12)
for pc in parsed_tuple[0]:
species_name = pc.metadata["species_name"]
self.assertTrue(species_name.startswith("spc"))
self.assertTrue(species_name[3:].isnumeric())
self.assertIsInstance(
pc, msprime.simulations.PopulationConfiguration)
self.assertIsInstance(parsed_tuple[1], list)
self.assertEqual(len(parsed_tuple[1]), 22)
event_types = [msprime.simulations.MassMigration]
event_types.append(msprime.simulations.PopulationParametersChange)
for mm in parsed_tuple[1]:
self.assertIn(type(mm), event_types)
def verify(
self,
tree,
pop_size_map,
nexus=None,
branch_length_units="yr",
generation_time=1,
):
if nexus is None:
nexus = make_nexus(tree, pop_size_map)
population_configurations, demographic_events = msprime.parse_starbeast(
nexus, generation_time, branch_length_units)
self.assertEqual(len(population_configurations), tree.num_samples())
for pop_config in population_configurations:
self.assertEqual(pop_config.growth_rate, 0)
self.assertIn("species_name", pop_config.metadata)
# Population IDs are mapped to leaves as they are encountered in a postorder
# traversal.
pop_id_map = {}
k = 0
for u in tree.nodes(order="postorder"):
if tree.is_leaf(u):
pop_id_map[u] = k
k += 1
else:
pop_id_map[u] = pop_id_map[tree.left_child(u)]
for u in tree.leaves():
pop_config = population_configurations[pop_id_map[u]]
self.assertEqual(pop_config.initial_size, pop_size_map[u])
self.assertEqual(pop_config.growth_rate, 0)
self.assertEqual(pop_config.metadata["species_name"], f"spc{u}")
# We should have demographic events for every non-unary internal node, and
# events should be output in increasing time order.
j = 0
for node in [u for u in tree.nodes(order="timeasc")]:
children = tree.children(node)
if len(children) > 1:
dest = pop_id_map[node]
for child in children[1:]:
event = demographic_events[j]
j += 1
self.assertIsInstance(event, msprime.MassMigration)
self.assertAlmostEqual(event.time, tree.time(node))
source = pop_id_map[child]
self.assertEqual(event.source, source)
self.assertEqual(event.dest, dest)
event = demographic_events[j]
j += 1
self.assertIsInstance(event,
msprime.PopulationParametersChange)
self.assertAlmostEqual(event.time, tree.time(node))
self.assertEqual(event.population, dest)
self.assertEqual(event.growth_rate, None)
self.assertEqual(event.initial_size, pop_size_map[node])
self.assertEqual(j, len(demographic_events))
def verify(
self,
tree,
pop_size_map,
nexus=None,
branch_length_units="yr",
generation_time=1,
):
if nexus is None:
nexus = make_nexus(tree, pop_size_map)
spec = msprime.parse_starbeast(nexus, generation_time,
branch_length_units)
assert spec.num_populations == tree.num_samples()
for pop in spec.populations:
assert pop.growth_rate == 0
# Population IDs are mapped to leaves as they are encountered in a postorder
# traversal.
pop_id_map = {}
k = 0
for u in tree.nodes(order="postorder"):
if tree.is_leaf(u):
pop_id_map[u] = k
k += 1
else:
pop_id_map[u] = pop_id_map[tree.left_child(u)]
for u in tree.leaves():
pop = spec.populations[pop_id_map[u]]
assert pop.initial_size == pop_size_map[u]
assert pop.growth_rate == 0
assert pop.name == f"spc{u}"
# We should have demographic events for every non-unary internal node, and
# events should be output in increasing time order.
j = 0
for node in [u for u in tree.nodes(order="timeasc")]:
children = tree.children(node)
if len(children) > 1:
dest = pop_id_map[node]
for child in children[1:]:
event = spec.events[j]
j += 1
assert isinstance(event, msprime.MassMigration)
self.assertAlmostEqual(event.time, tree.time(node))
source = pop_id_map[child]
assert event.source == source
assert event.dest == dest
event = spec.events[j]
j += 1
assert isinstance(event, msprime.PopulationParametersChange)
self.assertAlmostEqual(event.time, tree.time(node))
assert event.population == dest
assert event.growth_rate is None
assert event.initial_size == pop_size_map[node]
assert j == len(spec.events)
def test_bad_parameter(self):
with open("tests/data/species_trees/91genes_species_rev.tre") as f:
good_tree = f.read()
good_branch_length_units = "myr"
for bad_branch_length_units in [-3, "asdf", ["myr"], "gen"]:
with self.assertRaises(ValueError):
msprime.parse_starbeast(
tree=f.read(),
branch_length_units=bad_branch_length_units,
generation_time=5,
)
for bad_generation_time in [-3, "sdf"]:
with self.assertRaises(ValueError):
msprime.parse_starbeast(
tree=good_tree,
branch_length_units=good_branch_length_units,
generation_time=bad_generation_time,
)
for bad_generation_time in [None, {}]:
with self.assertRaises(TypeError):
msprime.parse_starbeast(
tree=good_tree,
branch_length_units=good_branch_length_units,
generation_time=bad_generation_time,
)
def test_12_species(self):
with open("tests/data/species_trees/91genes_species_rev.tre") as f:
good_tree = f.read()
good_branch_length_units = "myr"
good_generation_time = 5
spec = msprime.parse_starbeast(
tree=good_tree,
branch_length_units=good_branch_length_units,
generation_time=good_generation_time,
)
assert len(spec.populations) == 12
for pop in spec.populations:
species_name = pop.name
assert species_name.startswith("spc")
assert species_name[3:].isnumeric()
assert len(spec.events) == 22
event_types = [msprime.demography.MassMigration]
event_types.append(msprime.demography.PopulationParametersChange)
for mm in spec.events:
assert type(mm) in event_types