def test_EnforceTranslation_error_location_smaller_than_translation():
"""Providing a location that is not multiple of 3 raises an error"""
numpy.random.seed(1234)
sequence = reverse_translate(random_protein_sequence(15, seed=123))
with pytest.raises(ValueError) as err:
_ = DnaOptimizationProblem(
sequence=sequence,
constraints=[
EnforceTranslation(
translation=random_protein_sequence(30, seed=111))
],
logger=None,
)
assert str(err.value).startswith("Window size")
def experiment_1(seed=123):
"""A DNA chisel optimization whose results produced the file
test_determinism.py"""
np.random.seed(seed)
sequence = dc.reverse_translate(dc.random_protein_sequence(50))
# MAXIMIZE THE GC CONTENT
problem = dc.DnaOptimizationProblem(
sequence=sequence,
constraints=[dc.EnforceTranslation()],
objectives=[dc.EnforceGCContent(target=1)],
logger=None,
)
problem.optimize()
# BRING THE GC CONTENT BACK TO 50%
problem = dc.DnaOptimizationProblem(
sequence=problem.sequence,
constraints=[dc.EnforceTranslation()],
objectives=[dc.EnforceGCContent(target=0.5)],
logger=None,
)
problem.optimize()
return problem.sequence
def test_EnforceTranlationError():
"""Providing a location that is not multiple of 3 raises an error"""
numpy.random.seed(1234)
sequence = reverse_translate(random_protein_sequence(50, seed=123))
with pytest.raises(ValueError) as err:
problem = DnaOptimizationProblem(
sequence=sequence,
constraints=[EnforceTranslation(location=(0, 16))],
)
assert "Location 0-16(+) has length 16" in str(err.value)
def test_EnforceTranlation():
numpy.random.seed(1234)
sequence = reverse_translate(random_protein_sequence(50, seed=123))
problem = DnaOptimizationProblem(
sequence=sequence,
constraints=[AvoidPattern("AAA"), EnforceTranslation()],
)
assert not problem.all_constraints_pass()
problem.resolve_constraints()
assert problem.all_constraints_pass()
def test_codon_optimize_bestcodon():
numpy.random.seed(123)
protein = random_protein_sequence(3000, seed=123)
sequence = reverse_translate(protein)
problem = DnaOptimizationProblem(
sequence=sequence,
constraints=[EnforceTranslation()],
objectives=[CodonOptimize(species='e_coli')]
)
assert problem.objective_scores_sum() < 0
problem.optimize()
assert problem.objective_scores_sum() == 0
def test_codon_optimize_harmonized():
numpy.random.seed(123)
protein = random_protein_sequence(500, seed=123)
sequence = reverse_translate(protein)
problem = DnaOptimizationProblem(
sequence=sequence,
constraints=[EnforceTranslation()],
objectives=[CodonOptimize(species='e_coli', mode='harmonized')]
)
assert (-700 < problem.objective_scores_sum() < -600)
problem.optimize()
assert (-350 < problem.objective_scores_sum())
def test_maximal_protein_sequence_change():
np.random.seed(123)
protein = dc.random_protein_sequence(200)
sequence = dc.reverse_translate(protein)
problem = dc.DnaOptimizationProblem(
sequence=sequence,
constraints=[dc.EnforceTranslation()],
objectives=[dc.EnforceChanges()],
)
problem.resolve_constraints()
problem.optimize()
assert problem.number_of_edits() == 238
assert dc.translate(problem.sequence) == protein
def test_EnforceTranlationReversed():
numpy.random.seed(1234)
sequence = reverse_translate(random_protein_sequence(50, seed=123))
rev_sequence = reverse_complement(sequence)
problem = DnaOptimizationProblem(
sequence=rev_sequence,
constraints=[
AvoidPattern("AGC"),
EnforceTranslation(location=(0, len(sequence), -1))
],
)
assert not problem.all_constraints_pass()
problem.resolve_constraints()
assert problem.all_constraints_pass()
def experiment_2(seed=123):
np.random.seed(seed)
sequence = dc.reverse_translate(dc.random_protein_sequence(1000))
problem = dc.DnaOptimizationProblem(
sequence=sequence,
constraints=[
dc.EnforceTranslation(),
dc.EnforceGCContent(mini=0.4, maxi=0.6, window=50),
],
objectives=[dc.CodonOptimize(species="e_coli")],
logger=None,
)
problem.resolve_constraints()
problem.optimize()
return problem.sequence
def test_codon_optimize_harmonize_rca_short_sequence():
protein = random_protein_sequence(500, seed=123)
sequence = reverse_translate(protein)
harmonization = CodonOptimize(species="h_sapiens",
original_species="e_coli",
method="harmonize_rca")
problem = DnaOptimizationProblem(
sequence=sequence,
constraints=[EnforceTranslation()],
objectives=[harmonization],
logger=None,
)
assert problem.objective_scores_sum() < -123
problem.optimize()
assert -74 < problem.objective_scores_sum()
def test_codon_optimize_match_usage():
numpy.random.seed(123)
protein = random_protein_sequence(500, seed=123)
sequence = reverse_translate(protein)
problem = DnaOptimizationProblem(
sequence=sequence,
constraints=[EnforceTranslation()],
objectives=[
CodonOptimize(species="e_coli", method="match_codon_usage")
],
logger=None,
)
assert -600 < problem.objective_scores_sum() < -550
problem.optimize()
assert -350 < problem.objective_scores_sum()
def test_insert_and_erase_pattern():
numpy.random.seed(123)
protein = dc.random_protein_sequence(100)
pattern = "ATGC"
# CREATE A SEQUENCE WITH 0 PATTERN OCCURENCES
sequence = dc.random_compatible_dna_sequence(
sequence_length=300,
constraints=[
dc.EnforceTranslation(translation=protein),
dc.AvoidPattern(pattern),
],
logger=None,
)
# NOW INCREASE PATTERN OCCURENCES FROM 0 TO 5
problem = dc.DnaOptimizationProblem(
sequence=sequence,
constraints=[
dc.EnforcePatternOccurence(pattern, occurences=5),
dc.EnforceTranslation(),
],
logger=None,
)
assert problem.constraints[0].evaluate(problem).score == -5
problem.resolve_constraints()
assert problem.all_constraints_pass()
sequence = problem.sequence
# NOW DECREASE THE NUMBER OF OCCURENCES FROM 5 TO 2
problem = dc.DnaOptimizationProblem(
sequence=sequence,
constraints=[
dc.EnforcePatternOccurence(pattern, occurences=2),
dc.EnforceTranslation(),
],
logger=None,
)
assert problem.constraints[0].evaluate(problem).score == -3
problem.resolve_constraints()
assert problem.all_constraints_pass()
"""Example of use of the AvoidPAttern specification"""
from dnachisel import (
DnaOptimizationProblem,
random_protein_sequence,
reverse_translate,
CodonOptimize,
EnforceTranslation,
AvoidPattern,
EnforceGCContent,
)
protein = random_protein_sequence(1000, seed=123)
sequence = reverse_translate(protein)
problem = DnaOptimizationProblem(
sequence=sequence,
constraints=[
EnforceTranslation(),
AvoidPattern("BsmBI_site"),
EnforceGCContent(mini=0.4, maxi=0.6, window=60),
],
objectives=[CodonOptimize(species="s_cerevisiae")],
)
print("\nBefore optimization:\n")
print(problem.constraints_text_summary())
print(problem.objectives_text_summary())
problem.resolve_constraints(final_check=True)
problem.optimize()
all_9mers = [sequence[i:i + 9] for i in range(len(sequence) - 9)]
number_of_non_unique_9mers = sum([
count for ninemer, count in Counter(all_9mers).items() if count > 1
])
score = -(9.0 * number_of_non_unique_9mers) / len(sequence)
return SpecEvaluation(self,
problem,
score=score,
locations=[Location(0, len(sequence))],
message="Score: %.02f (%d non-unique ninemers)" %
(score, number_of_non_unique_9mers))
def __str__(self):
"""String representation."""
return "MinimizeNinemersScore"
sequence = reverse_translate(random_protein_sequence(300))
problem = DnaOptimizationProblem(sequence=sequence,
constraints=[EnforceTranslation()],
objectives=[MinimizeNinemersScore()])
print("\n=== Status before optimization ===")
print(problem.objectives_text_summary())
problem.optimize()
print("\n=== Status after optimization ===")
print(problem.objectives_text_summary())
print(problem.constraints_text_summary(failed_only=True))
