def test_get_raw_estimates(self):
"""correctly return raw result object"""
d = EstimateDistances(self.al, HKY85(), est_params=['kappa'])
d.run()
expect = {
('a', 'b'): {
'kappa': 1.0000226766004808e-06,
'length': 0.18232155856115662
},
('a', 'c'): {
'kappa': 1.0010380037049357e-06,
'length': 0.087070406623635604
},
('a', 'e'): {
'kappa': 2.3965871843412687,
'length': 0.4389176272584539
},
('b', 'e'): {
'kappa': 2.3965871854366592,
'length': 0.43891762729173389
},
('c', 'b'): {
'kappa': 1.0010380037049357e-06,
'length': 0.087070406623635604
},
('c', 'e'): {
'kappa': 0.57046787478038707,
'length': 0.43260232210282784
}
}
got = d.getAllParamValues()
for pair in expect:
for param in expect[pair]:
self.assertAlmostEqual(got[pair][param], expect[pair][param])
def test_EstimateDistances_other_model_params(self):
"""test getting other model params from EstimateDistances"""
d = EstimateDistances(self.al, HKY85(), est_params=['kappa'])
d.run()
# this will be a Number object with Mean, Median etc ..
kappa = d.getParamValues('kappa')
self.assertAlmostEqual(kappa.Mean, 0.8939, 4)
# this will be a dict with pairwise instances, it's called by the above
# method, so the correctness of it's values is already checked
kappa = d.getPairwiseParam('kappa')
def test_progressive_params(self):
"""excercise progressive alignment providing model params"""
self._test_aln(
{
'A': 'tacagta',
'B': 'tac-gtc',
'C': 'ta---ta',
'D': 'cac-cta',
},
model=HKY85(),
param_vals=[('kappa', 2.0)])
def test_EstimateDistancesWithMotifProbs(self):
"""EstimateDistances with supplied motif probs"""
motif_probs= {'A':0.1,'C':0.2,'G':0.2,'T':0.5}
d = EstimateDistances(self.al, HKY85(), motif_probs=motif_probs)
d.run()
canned_result = {('a', 'c'): 0.07537,
('b', 'c'): 0.07537,
('a', 'e'): 0.39921,
('a', 'b'): 0.15096,
('b', 'e'): 0.39921,
('c', 'e'): 0.37243}
result = d.getPairwiseDistances()
self.assertDistsAlmostEqual(canned_result, result)
def test_EstimateDistances_modify_lf(self):
"""tests modifying the lf"""
def constrain_fit(lf):
lf.setParamRule('kappa', is_const=True)
lf.optimise(local=True, show_progress=False)
return lf
d = EstimateDistances(self.al, HKY85(), modify_lf=constrain_fit)
d.run()
result = d.getPairwiseDistances()
d = EstimateDistances(self.al, F81())
d.run()
expect = d.getPairwiseDistances()
self.assertDistsAlmostEqual(expect, result)
def test_progressive_est_tree(self):
"""excercise progressive alignment without a guide tree"""
seqs = LoadSeqs(data={
'A':
"TGTGGCACAAATGCTCATGCCAGCTCTTTACAGCATGAGAACA",
'B':
"TGTGGCACAGATACTCATGCCAGCTCATTACAGCATGAGAACAGCAGTTT",
'C':
"TGTGGCACAAGTACTCATGCCAGCTCAGTACAGCATGAGAACAGCAGTTT"
},
aligned=False)
aln, tree = cogent.align.progressive.TreeAlign(
HKY85(), seqs, show_progress=False, param_vals={'kappa': 4.0})
expect = {
'A': 'TGTGGCACAAATGCTCATGCCAGCTCTTTACAGCATGAGAACA-------',
'C': 'TGTGGCACAAGTACTCATGCCAGCTCAGTACAGCATGAGAACAGCAGTTT',
'B': 'TGTGGCACAGATACTCATGCCAGCTCATTACAGCATGAGAACAGCAGTTT'
}
self.assertEqual(aln.todict(), expect)
def bh_fit(sa, st, return_lfs=None, global_optimisation=False, **kw):
standard_params = {'is_independent': True, 'is_constant': False}
optimise_params = {'local': not global_optimisation,
'show_progress': False, 'limit_action': 'raise'}
sm = HKY85()
lead_in_lf = sm.makeLikelihoodFunction(st)
lead_in_lf.setAlignment(sa)
lead_in_lf.optimise(**optimise_params)
sm = DiscreteSubstitutionModel(DNA.Alphabet, optimise_motif_probs=True,
model_gaps=False, recode_gaps=True,
name='DiscreteSubstitutionModel')
lf = sm.makeLikelihoodFunction(st)
lf.setAlignment(sa)
populate_parameters(lf, lead_in_lf, **standard_params)
lf.optimise(**optimise_params)
if return_lfs:
return [lf]
else:
return [deflate_likelihood_function(lf)]
#!/usr/bin/env python
# taken from http://pycogent.sourceforge.net/
from cogent import LoadSeqs, LoadTree
from cogent.evolve.models import HKY85
from cogent.maths import stats
aln = LoadSeqs(filename="data/long_testseqs.fasta")
t = LoadTree(filename="data/test.tree")
sm = HKY85()
lf = sm.makeLikelihoodFunction(t, digits=2, space=3)
lf.setLocalClock("Human", "HowlerMon")
lf.setAlignment(aln)
lf.optimise(show_progress=False)
lf.setName("clock")
print "lf =\n", lf
print "lf.getStatistics() =\n", lf.getStatistics()
opt_tr = lf.getAnnotatedTree()
print "opt_tr.getNewick() =\n", opt_tr.getNewick(with_distances=True)
null_lnL = lf.getLogLikelihood()
null_nfp = lf.getNumFreeParams()
lf.setParamRule('length', is_independent=True)
lf.optimise(show_progress=False)
lf.setName("non clock")
print lf
opt_tr = lf.getAnnotatedTree()
print "opt_tr.getNewick() =\n", opt_tr.getNewick(with_distances=True)
LR = 2 * (lf.getLogLikelihood() - null_lnL)
df = lf.getNumFreeParams() - null_nfp
P = stats.chisqprob(LR, df)
#!/usr/bin/env python
# taken from http://pycogent.sourceforge.net/
from cogent.evolve.models import HKY85
from cogent import LoadSeqs, LoadTree
model = HKY85()
aln = LoadSeqs("data/primate_cdx2_promoter.fasta")
tree = LoadTree(tip_names=aln.Names)
lf = model.makeLikelihoodFunction(tree)
lf.setAlignment(aln)
lf.optimise(show_progress = False)
print lf
