def test_order_0_states(self):
model_builder = hmm.pssm.ModelBuilder(2)
model = model_builder.create_background_mosaic_model(3, .01, 1.0)
positive = model_builder.add_order_0_parameterised_state(model, emission_dist=[.1,.2,.3,.4])
negative = model_builder.add_order_0_rev_comp_state(model, positive)
B = hmm.model_states_2_model(model).B[3:]
assert infpy.check_is_close_2(B[0,0], B[1,3])
assert infpy.check_is_close_2(B[0,1], B[1,2])
assert infpy.check_is_close_2(B[0,2], B[1,1])
assert infpy.check_is_close_2(B[0,3], B[1,0])
def test_order_0_states(self):
model_builder = hmm.pssm.ModelBuilder(2)
model = model_builder.create_background_mosaic_model(3, .01, 1.0)
positive = model_builder.add_order_0_parameterised_state(
model, emission_dist=[.1, .2, .3, .4])
negative = model_builder.add_order_0_rev_comp_state(model, positive)
B = hmm.model_states_2_model(model).B[3:]
assert infpy.check_is_close_2(B[0, 0], B[1, 3])
assert infpy.check_is_close_2(B[0, 1], B[1, 2])
assert infpy.check_is_close_2(B[0, 2], B[1, 1])
assert infpy.check_is_close_2(B[0, 3], B[1, 0])
def test_traits(self):
from hmm.pssm import create_background_model, seq_to_numpy
from infpy import check_is_close_2
for order in [0, 1, 2]:
num_background_mosaics = 4
traits = hmm.pssm.GappedPssmTraits(
K=7,
p_binding_site=.1,
background_order=order,
num_background_mosaics=num_background_mosaics,
background_model_creator=create_background_model
)
emission_dists = [
[ 1., 0., 0., 0. ],
[ 0., 1., 0., 0. ],
[ 0., 1., 0., 0. ],
[ 1., 0., 0., 0. ],
[ 0., 0., 1., 0. ],
[ 0., 0., 0., 1. ],
[ 0., 0., 0., 1. ],
[ 0., 0., 0., 1. ],
[ 0., 0., 1., 0. ],
[ 0., 1., 0., 0. ],
[ 1., 0., 0., 0. ],
[ 0., 1., 0., 0. ],
[ 0., 0., 0., 1. ],
]
m=hmm.as_model(traits.new_model(emission_dists))
# check we have parameterised the reverse complement gaps correctly
assert m.get_transition_parameterisation(29,28).idx == m.get_transition_parameterisation(14,15).idx
assert m.get_transition_parameterisation(29,28).idx != m.get_transition_parameterisation(4,5).idx
# check the reverse complement states are correct
B = m.B
for n in xrange(m.N):
for o in xrange(m.M):
rev_comp_state, rev_comp_obs = traits.get_non_reverse_complement(n,o)
assert check_is_close_2(
B[rev_comp_state,rev_comp_obs],
B[n,o]
), (
'%d,%d %d,%d: %f %f' % (
rev_comp_state,rev_comp_obs,n,o,B[rev_comp_state,rev_comp_obs],B[n,o]
)
)
# check viterbi gives correct result
test_seq = 'acgtgat' # matches dist above
test_seq_order_0 = hmm.pssm.seq_to_numpy(test_seq)
test_seq_order_n = m.converter.to_order_n(test_seq_order_0)
LL, states = m.viterbi(test_seq_order_n)
for i, state in enumerate(states):
assert (state-num_background_mosaics)/2 == i
def test_traits(self):
from hmm.pssm import create_background_model, PssmTraits, seq_to_numpy
from infpy import check_is_close_2
p_binding_site = .01
num_background_states = 2
emission_dists = [
[1., 0., 0., 0.],
[0., 1., 0., 0.],
[0., 1., 0., 0.],
[1., 0., 0., 0.],
[0., 0., 1., 0.],
[0., 0., 0., 1.],
[0., 0., 0., 1.],
[0., 0., 0., 1.],
[0., 0., 1., 0.],
[0., 1., 0., 0.],
[1., 0., 0., 0.],
[0., 1., 0., 0.],
[0., 0., 0., 1.],
]
K = len(emission_dists)
test_seq = 'accagtttgcact' # matches dist above
test_seq_order_0 = seq_to_numpy(test_seq)
# for various different orders
for order in [1, 2]:
# build a model of distribution above
traits = PssmTraits(K,
p_binding_site,
order,
num_background_states,
create_background_model,
emission_dists=emission_dists)
model = traits.new_model()
converted = hmm.model_states_2_model(model)
B = converted.B
# check the reverse complement states are correct
for n in xrange(model.N):
for o in xrange(model.M):
rev_comp_state, rev_comp_obs = traits.get_non_reverse_complement(
n, o)
assert check_is_close_2(
B[rev_comp_state, rev_comp_obs],
B[n, o]), ('%d,%d %d,%d: %f %f' %
(rev_comp_state, rev_comp_obs, n, o,
B[rev_comp_state, rev_comp_obs], B[n, o]))
# check viterbi gives correct result
test_seq_order_n = converted.converter.to_order_n(test_seq_order_0)
LL, states = converted.viterbi(test_seq_order_n)
for i, state in enumerate(states):
assert state == num_background_states + i
def test_traits(self):
from hmm.pssm import create_background_model, seq_to_numpy
from infpy import check_is_close_2
for order in [0, 1, 2]:
num_background_mosaics = 4
traits = hmm.pssm.GappedPssmTraits(
K=7,
p_binding_site=.1,
background_order=order,
num_background_mosaics=num_background_mosaics,
background_model_creator=create_background_model)
emission_dists = [
[1., 0., 0., 0.],
[0., 1., 0., 0.],
[0., 1., 0., 0.],
[1., 0., 0., 0.],
[0., 0., 1., 0.],
[0., 0., 0., 1.],
[0., 0., 0., 1.],
[0., 0., 0., 1.],
[0., 0., 1., 0.],
[0., 1., 0., 0.],
[1., 0., 0., 0.],
[0., 1., 0., 0.],
[0., 0., 0., 1.],
]
m = hmm.as_model(traits.new_model(emission_dists))
# check we have parameterised the reverse complement gaps correctly
assert m.get_transition_parameterisation(
29, 28).idx == m.get_transition_parameterisation(14, 15).idx
assert m.get_transition_parameterisation(
29, 28).idx != m.get_transition_parameterisation(4, 5).idx
# check the reverse complement states are correct
B = m.B
for n in xrange(m.N):
for o in xrange(m.M):
rev_comp_state, rev_comp_obs = traits.get_non_reverse_complement(
n, o)
assert check_is_close_2(
B[rev_comp_state, rev_comp_obs],
B[n, o]), ('%d,%d %d,%d: %f %f' %
(rev_comp_state, rev_comp_obs, n, o,
B[rev_comp_state, rev_comp_obs], B[n, o]))
# check viterbi gives correct result
test_seq = 'acgtgat' # matches dist above
test_seq_order_0 = hmm.pssm.seq_to_numpy(test_seq)
test_seq_order_n = m.converter.to_order_n(test_seq_order_0)
LL, states = m.viterbi(test_seq_order_n)
for i, state in enumerate(states):
assert (state - num_background_mosaics) / 2 == i
def test_traits(self):
from hmm.pssm import create_background_model, PssmTraits, seq_to_numpy
from infpy import check_is_close_2
p_binding_site = .01
num_background_states = 2
emission_dists = [
[ 1., 0., 0., 0. ],
[ 0., 1., 0., 0. ],
[ 0., 1., 0., 0. ],
[ 1., 0., 0., 0. ],
[ 0., 0., 1., 0. ],
[ 0., 0., 0., 1. ],
[ 0., 0., 0., 1. ],
[ 0., 0., 0., 1. ],
[ 0., 0., 1., 0. ],
[ 0., 1., 0., 0. ],
[ 1., 0., 0., 0. ],
[ 0., 1., 0., 0. ],
[ 0., 0., 0., 1. ],
]
K = len(emission_dists)
test_seq = 'accagtttgcact' # matches dist above
test_seq_order_0 = seq_to_numpy(test_seq)
# for various different orders
for order in [1, 2]:
# build a model of distribution above
traits = PssmTraits(K, p_binding_site, order, num_background_states, create_background_model, emission_dists=emission_dists)
model = traits.new_model()
converted = hmm.model_states_2_model(model)
B = converted.B
# check the reverse complement states are correct
for n in xrange(model.N):
for o in xrange(model.M):
rev_comp_state, rev_comp_obs = traits.get_non_reverse_complement(n,o)
assert check_is_close_2(B[rev_comp_state,rev_comp_obs], B[n,o]), ('%d,%d %d,%d: %f %f' % (rev_comp_state,rev_comp_obs,n,o,B[rev_comp_state,rev_comp_obs],B[n,o]))
# check viterbi gives correct result
test_seq_order_n = converted.converter.to_order_n(test_seq_order_0)
LL, states = converted.viterbi(test_seq_order_n)
for i, state in enumerate(states):
assert state == num_background_states+i
[ 0., 0., 1., 0. ],
[ 0., 1., 0., 0. ],
[ 1., 0., 0., 0. ],
[ 0., 1., 0., 0. ],
[ 0., 0., 0., 1. ],
]
K = len(emission_dists)
test_seq = 'accagtttgcact' # matches dist above
test_seq_order_0 = seq_to_numpy(test_seq)
# for various different orders
for order in [0, 1, 2]:
# build a model of distribution above
traits = PssmTraits(K, p_binding_site, order, num_background_states, create_background_model, emission_dists=emission_dists)
model = traits.new_model()
converted = hmm.model_states_2_model(model)
B = converted.B
# check the reverse complement states are correct
for n in xrange(model.N):
for o in xrange(model.M):
rev_comp_state, rev_comp_obs = traits.get_non_reverse_complement(n,o)
assert check_is_close_2(B[rev_comp_state,rev_comp_obs], B[n,o]), ('%d,%d %d,%d: %f %f' % (rev_comp_state,rev_comp_obs,n,o,B[rev_comp_state,rev_comp_obs],B[n,o]))
# check viterbi gives correct result
test_seq_order_n = converted.converter.to_order_n(test_seq_order_0)
LL, states = converted.viterbi(test_seq_order_n)
for i, state in enumerate(states):
assert state == num_background_states+i