def test_no_solution_error_exhaustive_search():
problem = DnaOptimizationProblem(
sequence="TTTTTTT",
constraints=[AvoidChanges((0, 4)), EnforceGCContent(mini=0.8)]
)
with pytest.raises(NoSolutionError) as err:
problem.resolve_constraints()
assert 'Exhaustive search failed' in str(err.value)
def test_no_solution_error_frozen_region():
problem = DnaOptimizationProblem(
sequence="AAAAATCGTCTCTTTT",
constraints=[AvoidChanges(), AvoidPattern('BsmBI_site')]
)
with pytest.raises(NoSolutionError) as err:
problem.resolve_constraints()
assert 'region that cannot be mutated' in str(err.value)
def test_no_solution_error_frozen_region():
problem = DnaOptimizationProblem(
sequence="AAAAATCGTCTCTTTT",
constraints=[AvoidChanges(),
AvoidPattern(enzyme='BsmBI')])
with pytest.raises(NoSolutionError) as err:
problem.resolve_constraints()
assert 'Constraint breach in frozen region' in str(err.value)
def test_no_solution_error_random_search():
problem = DnaOptimizationProblem(
sequence="TTTTTTTTTTTTTTTTTTTTTTTTTTTT",
constraints=[AvoidChanges((0, 10)), EnforceGCContent(mini=0.8)]
)
with pytest.raises(NoSolutionError) as err:
problem.resolve_constraints()
assert 'Random search did not' in str(err.value)
def test_pattern_and_reverse():
bsmbi = "CGTCTC"
bsmbi_rev = "GAGACG"
sequence = 10 * bsmbi + 25 * bsmbi_rev + 15 * bsmbi + 15 * bsmbi_rev
problem = DnaOptimizationProblem(sequence,
constraints=[AvoidPattern('BsmBI_site')],
objectives=[AvoidChanges()])
problem.resolve_constraints()
problem.optimize()
assert sum(problem.sequence_edits_as_array()) < 70
def test_optimization_with_report_no_solution(tmpdir):
problem = DnaOptimizationProblem(
sequence=random_dna_sequence(10000, seed=123),
constraints=[AvoidPattern(enzyme='BsmBI'), AvoidChanges()]
)
target = os.path.join(str(tmpdir), 'no_solution')
os.mkdir(target)
assert os.listdir(target) == []
success, message, data = optimization_with_report(target, problem)
assert not success
assert os.listdir(target) != []
def test_avoid_changes_with_indices_as_objectives():
numpy.random.seed(123)
indices = [10, 20] + list(range(30, 40)) + [44, 45, 46]
sequence = random_dna_sequence(50)
problem = DnaOptimizationProblem(
sequence=sequence,
objectives=[EnforceChanges(boost=0.5), AvoidChanges(indices=indices)],
logger=None,
)
problem.optimize()
assert problem.number_of_edits() == 50 - 15 # 15 == len(indices)
problem = DnaOptimizationProblem(
sequence=sequence,
objectives=[EnforceChanges(boost=1.5), AvoidChanges(indices=indices)],
logger=None,
)
problem.optimize()
assert problem.number_of_edits() == 50
def test_optimize_with_report_no_solution(tmpdir):
problem = DnaOptimizationProblem(
sequence=random_dna_sequence(10000, seed=123),
constraints=[AvoidPattern("BsmBI_site"), AvoidChanges()],
logger=None,
)
target = os.path.join(str(tmpdir), "no_solution")
os.mkdir(target)
assert os.listdir(target) == []
success, message, data = problem.optimize_with_report(target)
assert not success
assert os.listdir(target) != []
def test_AvoidChanges_with_max_edits():
numpy.random.seed(1)
problem = DnaOptimizationProblem(
sequence="ATATATATATA",
constraints=[
AvoidChanges(max_edits=2),
AvoidPattern("ATATA"),
EnforcePatternOccurence("A", occurences=6, location=(0, 11, 1)),
EnforcePatternOccurence("T", occurences=4, location=(0, 11, 1)),
],
logger=None,
)
assert not problem.all_constraints_pass()
problem.resolve_constraints()
assert problem.all_constraints_pass()
def test_avoid_change_as_objectives_basics():
numpy.random.seed(123)
results = []
for boost in (0, 0.1, 0.2, 1):
sequence = random_dna_sequence(1000, seed=123)
problem = DnaOptimizationProblem(
sequence=sequence,
objectives=[
EnforceGCContent(
mini=0.45, maxi=0.55,
window=80).copy_with_changes(locations_span=300),
AvoidChanges(boost=boost).as_passive_objective()
])
problem.optimize()
differences = sequences_differences(problem.sequence,
problem.sequence_before)
results.append(differences)
assert results[0] > 40
assert (results[0] > results[1] > results[2] > results[3])
assert results[-1] == 0
EnforceTranslation(),
EnforceGCContent(mini=0.35, maxi=0.65, window=50), #TWIST: 25% and 65% GC
],
objectives=[CodonOptimize(codon_usage_table=codon_table_11)],
)
if taxid and not input_path:
print("\nOptimizing codons for taxonomic ID: " + taxid)
#Read gene fasta sequence and initiate optimizer
if not protein_flag:
problem = DnaOptimizationProblem(
sequence=gene,
constraints=[
#EnforceSequence(sequence = "ATG", location=(0, 2)),
AvoidChanges(location=(0, 2)),
AvoidPattern("BsmBI_site", "BamHI"),
EnforceTranslation(),
EnforceGCContent(mini=0.35, maxi=0.65, window=50), #TWIST: 25% and 65% GC
],
objectives=[CodonOptimize(species=taxid)],
)
#Output and reporting
print("\nBefore optimization:")
print(problem.constraints_text_summary())
print(problem.objectives_text_summary())
problem.resolve_constraints(final_check=True)
def domesticate(
self,
dna_sequence=None,
protein_sequence=None,
is_cds="default",
codon_optimization=None,
extra_constraints=(),
extra_objectives=(),
final_record_target=None,
edit=False,
barcode="",
barcode_spacer="AA",
report_target=None,
):
"""Domesticate a sequence.
Parameters
----------
dna_sequence
The DNA sequence string to domesticate.
protein_sequence
Amino-acid sequence of the protein, which will be converted into
a DNA sequence string.
is_cds
If True, sequence edits are restricted to synonymous mutations.
codon_optimization
Either None for no codon optimization or the name of an organism
supported by DnaChisel.
extra_constraints
List of extra constraints to apply to the domesticated sequences.
Each constraint is either a DnaChisel constraint or a function
(dna_sequence => DnaChisel constraint).
extra_objectives
List of extra optimization objectives to apply to the domesticated
sequences. Each objective is either a DnaChisel constraint or a
function (dna_sequence => DnaChisel constraint).
final_record_target
Path to the file where to write the final genbank.
edit
Turn to True to allow sequence edits (if it is false and no all
constraints are originally satisfied, a failed domestication result
(i.e. with attribute ``success`` set to False) will be returned.
report_target
Target for the sequence optimization report (a folder path, or a zip
path).
barcode
A sequence of DNA that will be added to the left of the sequence once
the domestication is done.
barcode_spacer
Nucleotides to be added between the barcode and the enzyme (optional,
the idea here is that they will make sure to avoid the creation of
unwanted cutting sites).
Returns
-------
final_record, edits_record, report_data, success, msg
"""
if is_cds == "default":
is_cds = self.cds_by_default
if isinstance(dna_sequence, SeqRecord):
problem = DnaOptimizationProblem.from_record(dna_sequence)
for spec in problem.constraints + problem.objectives:
spec.location += len(self.left_flank)
extra_constraints = list(extra_constraints) + problem.constraints
extra_objectives = list(extra_constraints) + problem.objectives
if protein_sequence is not None:
is_cds = True
dna_sequence = reverse_translate(protein_sequence)
constraints = [
c(dna_sequence) if hasattr(c, "__call__") else c
for c in list(extra_constraints) + self.constraints
]
location = Location(len(self.left_flank),
len(self.left_flank) + len(dna_sequence))
if is_cds:
constraints.append(EnforceTranslation(location=location))
objectives = [
o(dna_sequence) if hasattr(o, "__call__") else o
for o in list(extra_objectives) + self.objectives
]
if codon_optimization:
objectives.append(
CodonOptimize(species=codon_optimization, location=location))
if self.minimize_edits:
objectives.append(AvoidChanges())
extended_sequence = self.left_flank + dna_sequence + self.right_flank
if (not is_cds) and (not edit):
constraints.append(AvoidChanges())
problem = DnaOptimizationProblem(
extended_sequence,
constraints=constraints,
objectives=objectives,
logger=self.logger,
)
all_constraints_pass = problem.all_constraints_pass()
no_objectives = (len(problem.objectives) - self.minimize_edits) == 0
report_data = None
optimization_successful = True
message = ""
# print (all_constraints_pass, no_objectives)
if not (all_constraints_pass and no_objectives):
problem.n_mutations = self.simultaneous_mutations
if report_target is not None:
(success, message, report_data) = problem.optimize_with_report(
target=report_target, project_name=self.name)
optimization_successful = success
else:
report_data = None
try:
problem.resolve_constraints()
problem.optimize()
except Exception as err:
message = str(err)
optimization_successful = False
report_data = None
final_record = problem.to_record(
with_original_features=True,
with_original_spec_features=False,
with_constraints=False,
with_objectives=False,
)
edits_record = problem.to_record(
with_original_features=True,
with_original_spec_features=False,
with_constraints=False,
with_objectives=False,
with_sequence_edits=True,
)
if final_record_target is not None:
SeqIO.write(final_record, final_record_target, "genbank")
return DomesticationResult(
problem.sequence_before,
final_record,
edits_record,
report_data,
optimization_successful,
message,
)
AvoidPattern("9xA"),
AvoidPattern("9xT"),
AvoidPattern(HomopolymerPattern("6xG")),
AvoidPattern(HomopolymerPattern("6xC")),
EnforceGCContent(0.4, 0.65),
EnforceGCContent(0.25, 0.80, window=50),
]
CDS_constraints = []
for (start, end, strand) in CDS_list:
if strand == 1:
promoter_region = (start - 30, start - 1)
else:
promoter_region = (end + 1, end + 30)
CDS_constraints += [
AvoidChanges(promoter_region),
EnforceTranslation((start, end, strand)),
]
# DEFINE OBJECTIVES
objectives = [EnforceGCContent(0.51, boost=10000)] + [
CodonOptimize("e_coli", location=(start, end, strand))
for (start, end, strand) in CDS_list
]
# DEFINE AND SOLVE THE PROBLEM
problem = DnaOptimizationProblem(
sequence=record,
constraints=dna_provider_constraints + CDS_constraints,
"""Example of use of the AvoidChanges as an objective to minimize modifications
of a sequence."""
from dnachisel import (DnaOptimizationProblem, random_dna_sequence,
AvoidPattern, AvoidChanges, sequences_differences,
EnforceGCContent)
# Note: we are not providing a location for AvoidChanges: it applies globally
for boost in (0, 0.1, 1, 10.0):
sequence = random_dna_sequence(1000, seed=123)
problem = DnaOptimizationProblem(
sequence=sequence,
objectives=[
EnforceGCContent(mini=0.45, maxi=0.55, window=80),
AvoidChanges(boost=boost).as_passive_objective()
])
problem.optimize()
differences = sequences_differences(problem.sequence,
problem.sequence_before)
print("%d nucleotides modified for boost=%.1f" % (differences, boost))
