def test_number_of_calls(self):
# Test that genetic map citations are converted.
species = stdpopsim.get_species("HomSap")
genetic_map = species.get_genetic_map("HapMapII_GRCh37")
contig = species.get_contig("chr22", genetic_map=genetic_map.id)
model = stdpopsim.PiecewiseConstantSize(species.population_size)
engine = stdpopsim.get_default_engine()
cites_and_cites = [
genetic_map.citations,
model.citations,
engine.citations,
species.genome.mutation_rate_citations,
species.genome.recombination_rate_citations,
species.genome.assembly_citations,
]
ncite = len(set([ref.doi for cites in cites_and_cites for ref in cites]))
# Patch out writing to a file, then
# ensure that the method is called
# the correct number of times.
with mock.patch("builtins.open", mock.mock_open()):
with open('tmp.bib', 'w') as bib:
with mock.patch.object(
stdpopsim.citations.Citation,
"fetch_bibtex") as mock_bib:
cli.write_bibtex(engine, model, contig, species, bib)
self.assertEqual(mock_bib.call_count, ncite)
def test_number_of_calls(self):
# Test that genetic map citations are converted.
species = stdpopsim.get_species("HomSap")
genetic_map = species.get_genetic_map("HapMapII_GRCh37")
contig = species.get_contig("chr20", genetic_map=genetic_map.id)
model = stdpopsim.PiecewiseConstantSize(species.population_size)
engine = stdpopsim.get_default_engine()
local_cites = stdpopsim.Citation.merge(
[stdpopsim.citations._stdpopsim_citation]
+ genetic_map.citations
+ model.citations
+ engine.citations
+ species.genome.citations
+ species.citations
)
dois = set([ref.doi for ref in local_cites])
ncite = len(dois)
assert ncite == len(local_cites)
cli_cites = cli.get_citations(engine, model, contig, species)
assert len(cli_cites) == len(local_cites)
# Patch out writing to a file, then
# ensure that the method is called
# the correct number of times.
with mock.patch("builtins.open", mock.mock_open()):
with open("tmp.bib", "w") as bib:
with mock.patch.object(
stdpopsim.citations.Citation, "fetch_bibtex", autospec=True
) as mock_bib:
cli.write_bibtex(engine, model, contig, species, bib)
assert mock_bib.call_count == ncite
def generic_models_example():
species = stdpopsim.get_species("HomSap")
contig = species.get_contig("chr22", length_multiplier=0.1)
model = stdpopsim.PiecewiseConstantSize(species.population_size)
samples = model.get_samples(10)
engine = stdpopsim.get_default_engine()
ts = engine.simulate(model, contig, samples)
def test_genetic_map_citations(self):
species = stdpopsim.get_species("HomSap")
genetic_map = species.get_genetic_map("HapMapII_GRCh37")
contig = species.get_contig("chr22", genetic_map=genetic_map.id)
model = stdpopsim.PiecewiseConstantSize(species.population_size)
engine = stdpopsim.get_default_engine()
stdout, stderr = capture_output(
cli.write_citations, engine, model, contig, species)
self.assertEqual(len(stdout), 0)
self.check_citations(engine, species, genetic_map, model, stderr)
def test_model_citations(self):
contig = stdpopsim.Contig()
species = stdpopsim.get_species("HomSap")
model = species.get_demographic_model("OutOfAfrica_3G09")
engine = stdpopsim.get_default_engine()
stdout, stderr = capture_output(
cli.write_citations, engine, model, contig, species)
self.assertEqual(len(stdout), 0)
genetic_map = None
self.check_citations(engine, species, genetic_map, model, stderr)
def test_model_citations(self, caplog):
species = stdpopsim.get_species("HomSap")
contig = species.get_contig("22")
model = species.get_demographic_model("OutOfAfrica_3G09")
engine = stdpopsim.get_default_engine()
stdout, stderr = capture_output(
cli.write_citations, engine, model, contig, species
)
assert len(stdout) == 0
genetic_map = None
self.check_citations(engine, species, genetic_map, model, caplog.text)
def test_genetic_map_citations(self, caplog):
species = stdpopsim.get_species("HomSap")
genetic_map = species.get_genetic_map("HapMapII_GRCh37")
contig = species.get_contig("chr20", genetic_map=genetic_map.id)
model = stdpopsim.PiecewiseConstantSize(species.population_size)
engine = stdpopsim.get_default_engine()
dfe = None
stdout, stderr = capture_output(cli.write_citations, engine, model,
contig, species, dfe)
assert len(stdout) == 0
self.check_citations(engine, species, genetic_map, model, caplog.text)
def test_dfe_citations(self, caplog):
species = stdpopsim.get_species("HomSap")
genetic_map = species.get_genetic_map("HapMapII_GRCh37")
dfe = species.get_genetic_map("HapMapII_GRCh37")
contig = species.get_contig("chr20", genetic_map=genetic_map.id)
model = stdpopsim.PiecewiseConstantSize(species.population_size)
engine = stdpopsim.get_default_engine()
dfe = species.get_dfe("Gamma_K17")
stdout, stderr = capture_output(cli.write_citations, engine, model,
contig, species, dfe)
assert len(stdout) == 0
assert "[distribution of fitness effects]" in caplog.text
assert "Kim et al., 2017" in caplog.text
def test_simulation_runs(self):
# With a recombination_map of None, we simulate a coalescent without
# recombination in msprime, with no mutation.
contig = stdpopsim.Contig()
# Generate vector with 2 samples for each pop with sampling enabled
sample_count = []
for p in self.model.populations:
if p.allow_samples:
sample_count.append(2)
else:
sample_count.append(0)
samples = self.model.get_samples(*sample_count)
engine = stdpopsim.get_default_engine()
ts = engine.simulate(self.model, contig, samples)
self.assertEqual(ts.num_populations, self.model.num_populations)
def test_simulation_runs(self):
# With a recombination_map of None, we simulate a coalescent without
# recombination in msprime, with mutation rate equal to rate from model.
contig = stdpopsim.Contig.basic_contig(
length=100, mutation_rate=self.model.mutation_rate
)
# Generate vector with 2 samples for each pop with sampling enabled
sample_count = []
for p in self.model.populations:
if p.allow_samples:
sample_count.append(2)
else:
sample_count.append(0)
samples = self.model.get_samples(*sample_count)
engine = stdpopsim.get_default_engine()
ts = engine.simulate(self.model, contig, samples)
assert ts.num_populations == self.model.num_populations
def run_chr20_ooa(samples,
Ne,
length,
mutation_rate,
recombination_rate,
rng,
seed=None):
"""
Run StandardPopSim Out of Africa Chromosome 20
"""
species = stdpopsim.get_species("HomSap")
contig = species.get_contig("chr20", genetic_map="HapMapII_GRCh37")
model = species.get_demographic_model("OutOfAfrica_3G09")
engine = stdpopsim.get_default_engine()
ts = engine.simulate(model, contig, samples, seed=seed)
snippet_start = rng.randint(0, ts.get_sequence_length() - length)
snippet = [snippet_start, snippet_start + length]
return ts.keep_intervals(np.array([snippet])).trim()
def test_register_duplicate(self):
engine = stdpopsim.get_default_engine()
with self.assertRaises(ValueError):
stdpopsim.register_engine(engine)
def test_get_default_engine(self):
engine = stdpopsim.get_default_engine()
self._test_engine(engine)
def stdpopsim_cli_parser():
# TODO the CLI defined by this hierarchical and clumsy, but it's the best
# I could figure out. It can definitely be improved!
top_parser = argparse.ArgumentParser(
description="Command line interface for stdpopsim.")
top_parser.add_argument("-V",
"--version",
action='version',
version='%(prog)s {}'.format(
stdpopsim.__version__))
top_parser.add_argument("-v",
"--verbosity",
action='count',
default=0,
help="Increase the verbosity")
top_parser.add_argument(
"-c",
"--cache-dir",
type=str,
default=None,
help=("Set the cache directory to the specified value. "
"Note that this can also be set using the environment variable "
"STDPOPSIM_CACHE. If both the environment variable and this "
"option are set, the option takes precedence. "
f"Default: {stdpopsim.get_cache_dir()}"))
top_parser.add_argument("-e",
"--engine",
default=stdpopsim.get_default_engine().id,
choices=[e.id for e in stdpopsim.all_engines()],
help="Specify a simulation engine.")
for engine in stdpopsim.all_engines():
group = top_parser.add_argument_group(
f"{engine.id} specific parameters")
engine.add_arguments(group)
subparsers = top_parser.add_subparsers(dest="subcommand")
subparsers.required = True
for species in stdpopsim.all_species():
add_simulate_species_parser(subparsers, species)
download_maps_parser = subparsers.add_parser(
"download-genetic-maps",
help="Download genetic maps",
description=(
"Download genetic maps and store them in the cache directory. "
"Maps are downloaded regardless of whether they are already "
"in the cache or not. Please use the --cache-dir option to "
"download maps to a specific directory. "))
download_maps_parser.add_argument(
"species",
nargs="?",
help=("Download genetic maps for this species. If not specified "
"download all known genetic maps."))
download_maps_parser.add_argument(
"genetic_maps",
type=str,
nargs="*",
help=("If specified, download these genetic maps. If no maps "
"are provided, download all maps for this species."))
download_maps_parser.set_defaults(runner=run_download_genetic_maps)
return top_parser
def stdpopsim_cli_parser():
# TODO the CLI defined by this hierarchical and clumsy, but it's the best
# I could figure out. It can definitely be improved!
top_parser = argparse.ArgumentParser(
description="Command line interface for stdpopsim.")
top_parser.add_argument("-V",
"--version",
action='version',
version='%(prog)s {}'.format(
stdpopsim.__version__))
top_parser.add_argument("-v",
"--verbosity",
action='count',
default=1,
help="Increase the verbosity")
top_parser.add_argument(
"-c",
"--cache-dir",
type=str,
default=None,
help=("Set the cache directory to the specified value. "
"Note that this can also be set using the environment variable "
"STDPOPSIM_CACHE. If both the environment variable and this "
"option are set, the option takes precedence. "
f"Default: {stdpopsim.get_cache_dir()}"))
top_parser.add_argument("-e",
"--engine",
default=stdpopsim.get_default_engine().id,
choices=[e.id for e in stdpopsim.all_engines()],
help="Specify a simulation engine.")
supported_models = stdpopsim.get_engine("msprime").supported_models
msprime_parser = top_parser.add_argument_group(
"msprime specific parameters")
msprime_parser.add_argument(
"--msprime-model",
default=supported_models[0],
choices=supported_models,
help="Specify the simulation model used by msprime. "
"See msprime API documentation for details.")
def time_or_model(arg, _arg_is_time=[
True,
], parser=top_parser):
if _arg_is_time[0]:
try:
arg = float(arg)
except ValueError:
parser.error(f"`{arg}' is not a number")
else:
if arg not in supported_models:
parser.error(f"`{arg}' is not a supported model")
_arg_is_time[0] = not _arg_is_time[0]
return arg
msprime_parser.add_argument(
"--msprime-change-model",
metavar=("T", "MODEL"),
type=time_or_model,
default=[],
action="append",
nargs=2,
help="Change to the specified simulation MODEL at generation T. "
"This option may provided multiple times.")
# SLiM is not available for windows.
if not IS_WINDOWS:
def slim_exec(path):
# Hack to set the SLIM environment variable at parse time,
# before get_version() can be called.
os.environ["SLIM"] = path
return path
slim_parser = top_parser.add_argument_group("SLiM specific parameters")
slim_parser.add_argument("--slim-path",
metavar="PATH",
type=slim_exec,
default=None,
help="Full path to `slim' executable.")
slim_parser.add_argument(
"--slim-script",
action="store_true",
default=False,
help="Write script to stdout and exit without running SLiM.")
slim_parser.add_argument(
"--slim-scaling-factor",
metavar="Q",
default=1,
type=float,
help="Rescale model parameters by Q to speed up simulation. "
"See SLiM manual: `5.5 Rescaling population sizes to "
"improve simulation performance`. "
"[default=%(default)s].")
slim_parser.add_argument(
"--slim-burn-in",
metavar="X",
default=10,
type=float,
help="Length of the burn-in phase, in units of N generations "
"[default=%(default)s].")
subparsers = top_parser.add_subparsers(dest="subcommand")
subparsers.required = True
for species in stdpopsim.all_species():
add_simulate_species_parser(subparsers, species)
download_maps_parser = subparsers.add_parser(
"download-genetic-maps",
help="Download genetic maps",
description=(
"Download genetic maps and store them in the cache directory. "
"Maps are downloaded regardless of whether they are already "
"in the cache or not. Please use the --cache-dir option to "
"download maps to a specific directory. "))
download_maps_parser.add_argument(
"species",
nargs="?",
help=("Download genetic maps for this species. If not specified "
"download all known genetic maps."))
download_maps_parser.add_argument(
"genetic_maps",
type=str,
nargs="*",
help=("If specified, download these genetic maps. If no maps "
"are provided, download all maps for this species."))
download_maps_parser.set_defaults(runner=run_download_genetic_maps)
return top_parser
species = stdpopsim.get_species("DroMel")
## I have specified the desired chromosome arm at the command line, let's
contig = species.get_contig(sys.argv[1], genetic_map = "ComeronCrossover_dm6")
## For testing, it is good to model a lil chunk of chromosome
#contig = species.get_contig(sys.argv[1], length_multiplier = 0.10)
## You can grab the genetic map out of the simulations using:
#for p, r in zip( contig.recombination_map.get_positions() , contig.recombination_map.get_rates() ):
# print( p , r )
Ne = species.population_size/10
#Ne = 10000
model = stdpopsim.IsolationWithMigration(2*Ne, Ne, Ne, Ne, 1.5/Ne, 1.5/Ne)
print("NA", "N1", "N2", "T", "M12", "M21")
print( 2*Ne, Ne, Ne, Ne, 1.5/Ne, 1.5/Ne)
## I want to simulate 20 samples from each population
samples = model.get_samples(20,20)
## I will simulate using msprime
engine = stdpopsim.get_default_engine()
print("running simulation")
ts = engine.simulate(model, contig, samples)
## Save the simulated data to a VCF
with open("drosophilaSimulated."+ sys.argv[1] + ".vcf", "w") as vcf_file:
ts.write_vcf(vcf_file)
