def test_sample_arg_smc(self):
"""Sample an ARG using the SMC process"""
length = 10000 # length of locus
k = 5 # number of lineages
n = 1e4 # effective popsize
rho = 1.5e-8 # recomb/site/gen
arg = arglib.sample_arg_smc(k, n, rho, 0, length)
arglib.assert_arg(arg)
def test_est_popsize2(self):
"""
Fully sample an ARG from stratch using API
"""
k = 20
rho = 1.5e-8
mu = 2.5e-8
length = int(4e6)
popsize = 1e4
popsize2 = 1e4 * .5
a = int(.3 * length)
b = int(.7 * length)
refine = 0
util.tic("sim ARG")
arg = arglib.sample_arg_smc(k, 2 * popsize,
rho, start=0, end=a)
arg = arglib.sample_arg_smc(k, 2 * popsize2,
rho, start=a, end=b,
init_tree=arg)
arg = arglib.sample_arg_smc(k, 2 * popsize,
rho, start=b, end=length,
init_tree=arg)
util.toc()
x = []; y = []
for (start, end), tree in arglib.iter_tree_tracks(arg):
arglib.remove_single_lineages(tree)
x.append(start)
y.append(mle_popsize_tree(tree, mintime=0))
x2, y2 = stats.smooth2(x, y, 100e3)
p = plot(x, y, ymin=0)
p.plot(x2, y2, style='lines')
p.plot([0, a, a, b, b, length],
[popsize, popsize, popsize2, popsize2, popsize, popsize],
style='lines')
pause()
def test_est_popsize2(self):
"""
Fully sample an ARG from stratch using API
"""
k = 20
rho = 1.5e-8
mu = 2.5e-8
length = int(4e6)
popsize = 1e4
popsize2 = 1e4 * .5
a = int(.3 * length)
b = int(.7 * length)
refine = 0
util.tic("sim ARG")
arg = arglib.sample_arg_smc(k, 2 * popsize,
rho, start=0, end=a)
arg = arglib.sample_arg_smc(k, 2 * popsize2,
rho, start=a, end=b,
init_tree=arg)
arg = arglib.sample_arg_smc(k, 2 * popsize,
rho, start=b, end=length,
init_tree=arg)
util.toc()
x = []; y = []
for (start, end), tree in arglib.iter_tree_tracks(arg):
arglib.remove_single_lineages(tree)
x.append(start)
y.append(mle_popsize_tree(tree, mintime=0))
x2, y2 = stats.smooth2(x, y, 100e3)
p = plot(x, y, ymin=0)
p.plot(x2, y2, style='lines')
p.plot([0, a, a, b, b, length],
[popsize, popsize, popsize2, popsize2, popsize, popsize],
style='lines')
pause()
def test_forward():
k = 4
n = 1e4
rho = 1.5e-8 * 20
mu = 2.5e-8 * 20
length = int(100e3 / 20)
times = argweaver.get_time_points(ntimes=100)
arg = arglib.sample_arg_smc(k, 2 * n, rho, start=0, end=length)
muts = arglib.sample_arg_mutations(arg, mu)
seqs = arglib.make_alignment(arg, muts)
print "muts", len(muts)
print "recomb", len(arglib.get_recomb_pos(arg))
argweaver.discretize_arg(arg, times)
# remove chrom
new_name = "n%d" % (k - 1)
arg = argweaver.remove_arg_thread(arg, new_name)
carg = argweaverc.arg2ctrees(arg, times)
util.tic("C fast")
probs1 = argweaverc.argweaver_forward_algorithm(carg, seqs, times=times)
util.toc()
util.tic("C slow")
probs2 = argweaverc.argweaver_forward_algorithm(carg,
seqs,
times=times,
slow=True)
util.toc()
for i, (col1, col2) in enumerate(izip(probs1, probs2)):
for a, b in izip(col1, col2):
fequal(a, b, rel=.0001)
def test_forward():
k = 4
n = 1e4
rho = 1.5e-8 * 20
mu = 2.5e-8 * 20
length = int(100e3 / 20)
times = argweaver.get_time_points(ntimes=100)
arg = arglib.sample_arg_smc(k, 2*n, rho, start=0, end=length)
muts = arglib.sample_arg_mutations(arg, mu)
seqs = arglib.make_alignment(arg, muts)
print "muts", len(muts)
print "recomb", len(arglib.get_recombs(arg))
argweaver.discretize_arg(arg, times)
# remove chrom
new_name = "n%d" % (k - 1)
arg = argweaver.remove_arg_thread(arg, new_name)
carg = argweaverc.arg2ctrees(arg, times)
util.tic("C fast")
probs1 = argweaverc.argweaver_forward_algorithm(carg, seqs, times=times)
util.toc()
util.tic("C slow")
probs2 = argweaverc.argweaver_forward_algorithm(carg, seqs, times=times,
slow=True)
util.toc()
for i, (col1, col2) in enumerate(izip(probs1, probs2)):
for a, b in izip(col1, col2):
fequal(a, b, rel=.0001)
smc_coals_list = []
for i in range(20):
k = 10
n = 10e3
length = 500e3
rho = 1.5e-8
# simulate an ARG from the CwR and convert it into SMC-style
tic("simulate %d" % i)
cwr_arg = arglib.sample_arg(k, n, rho, start=0, end=length)
cwr_arg_converted = arglib.smcify_arg(cwr_arg)
toc()
# simulate an ARG directly from SMC process
smc_arg = arglib.sample_arg_smc(k, n, rho, start=0, end=length)
# gather all coalescence times
cwr_coals = [node.age for node in cwr_arg_converted
if node.event == 'coal']
smc_coals = [node.age for node in smc_arg
if node.event == 'coal']
print len(cwr_coals), len(smc_coals)
cwr_coals_list.append(cwr_coals)
smc_coals_list.append(smc_coals)
rplot_start('figures/cwr-smc-coals.pdf')
rp.plot([], main='Comparison of CwR and SMC coalescence times',
xlab='generations', ylab='', xlim=[50, 100e3], ylim=[0, 1],
def test_est_arg_popsize(self):
"""
Fully sample an ARG from stratch using API
"""
k = 20
rho = 1.5e-8 * 20
mu = 2.5e-8 * 20
length = int(2e6) / 20
times = arghmm.get_time_points(ntimes=20, maxtime=200000)
popsize = 1e4
popsize2 = 1e4 * .5
a = int(.3 * length)
b = int(.7 * length)
refine = 0
util.tic("sim ARG")
arg = arglib.sample_arg_smc(k, 2 * popsize,
rho, start=0, end=a)
arg = arglib.sample_arg_smc(k, 2 * popsize2,
rho, start=a, end=b,
init_tree=arg)
arg = arglib.sample_arg_smc(k, 2 * popsize,
rho, start=b, end=length,
init_tree=arg)
# sim seq
mut = arghmm.sample_arg_mutations(arg, mu, times)
seqs = arghmm.make_alignment(arg, mut)
util.toc()
# sample arg
util.tic("sample arg")
arg2 = arghmm.sample_arg(seqs, rho=rho, mu=mu, times=times,
popsizes=1e4, carg=True)
arg2 = arghmm.resample_climb_arg(arg2, seqs, popsizes=1e4,
rho=rho, mu=mu, times=times,
refine=200)
arg2 = arghmm.resample_all_arg(arg2, seqs, popsizes=1e4,
rho=rho, mu=mu, times=times,
refine=200)
util.toc()
x = []; y = []
for (start, end), tree in arglib.iter_tree_tracks(arg2):
arglib.remove_single_lineages(tree)
x.append(start)
y.append(mle_popsize_tree(tree, mintime=0))
# thin popsizes
x2 = list(range(0, length, length//5000)); y2 = []
j = 0
for i in range(len(x2)):
while j < len(x) and x[j] < x2[i]:
j += 1
y2.append(y[min(j, len(y)-1)])
x3, y3 = stats.smooth2(x2, y2, 100e3)
p = plot(x, y, ymin=0)
p.plot(x3, y3, style='lines')
p.plot([0, a, a, b, b, length],
[popsize, popsize, popsize2, popsize2, popsize, popsize],
style='lines')
pause()
def test_est_arg_popsize(self):
"""
Fully sample an ARG from stratch using API
"""
k = 20
rho = 1.5e-8 * 20
mu = 2.5e-8 * 20
length = int(2e6) / 20
times = arghmm.get_time_points(ntimes=20, maxtime=200000)
popsize = 1e4
popsize2 = 1e4 * .5
a = int(.3 * length)
b = int(.7 * length)
refine = 0
util.tic("sim ARG")
arg = arglib.sample_arg_smc(k, 2 * popsize,
rho, start=0, end=a)
arg = arglib.sample_arg_smc(k, 2 * popsize2,
rho, start=a, end=b,
init_tree=arg)
arg = arglib.sample_arg_smc(k, 2 * popsize,
rho, start=b, end=length,
init_tree=arg)
# sim seq
mut = arghmm.sample_arg_mutations(arg, mu, times)
seqs = arghmm.make_alignment(arg, mut)
util.toc()
# sample arg
util.tic("sample arg")
arg2 = arghmm.sample_arg(seqs, rho=rho, mu=mu, times=times,
popsizes=1e4, carg=True)
arg2 = arghmm.resample_climb_arg(arg2, seqs, popsizes=1e4,
rho=rho, mu=mu, times=times,
refine=200)
arg2 = arghmm.resample_all_arg(arg2, seqs, popsizes=1e4,
rho=rho, mu=mu, times=times,
refine=200)
util.toc()
x = []; y = []
for (start, end), tree in arglib.iter_tree_tracks(arg2):
arglib.remove_single_lineages(tree)
x.append(start)
y.append(mle_popsize_tree(tree, mintime=0))
# thin popsizes
x2 = range(0, length, length//5000); y2 = []
j = 0
for i in range(len(x2)):
while j < len(x) and x[j] < x2[i]:
j += 1
y2.append(y[min(j, len(y)-1)])
x3, y3 = stats.smooth2(x2, y2, 100e3)
p = plot(x, y, ymin=0)
p.plot(x3, y3, style='lines')
p.plot([0, a, a, b, b, length],
[popsize, popsize, popsize2, popsize2, popsize, popsize],
style='lines')
pause()
