def get_activation(sequences, concentrations, params):
model = nu.Model(material='dna',
celsius=params['temperature'],
sodium=params['sodium'],
magnesium=params['magnesium'])
# Define strands and sequences
probeF = nu.Strand(sequences['probeF'], name='probeF')
probeQ = nu.Strand(sequences['probeQ'], name='probeQ')
sink = nu.Strand(sequences['sink'], name='sink')
sinkC = nu.Strand(sequences['sinkC'], name='sinkC')
non_mut_target = nu.Strand(sequences['non_mut_target'],
name='non_mut_target')
mut_target = nu.Strand(sequences['mut_target'], name='mut_target')
# Define 3 tubes: one without target (initial), one with mutated target (mut), and one with non-mutated target (non_mut)
initial_state = nu.Tube(strands={
probeF: concentrations['probeF'],
probeQ: concentrations['probeQ'],
sink: concentrations['sink'],
sinkC: concentrations['sinkC']
},
complexes=nu.SetSpec(
max_size=2,
include=[[probeF, probeQ, sink, sinkC]]),
name='initial_state')
mut_state = nu.Tube(strands={
probeF: concentrations['probeF'],
probeQ: concentrations['probeQ'],
sink: concentrations['sink'],
sinkC: concentrations['sinkC'],
mut_target: concentrations['mut_target']
},
complexes=nu.SetSpec(
max_size=2,
include=[[probeF, probeQ, sink, sinkC,
mut_target]]),
name='mut_state')
non_mut_state = nu.Tube(strands={
probeF: concentrations['probeF'],
probeQ: concentrations['probeQ'],
sink: concentrations['sink'],
sinkC: concentrations['sinkC'],
non_mut_target: concentrations['non_mut_target']
},
complexes=nu.SetSpec(max_size=2,
include=[[
probeF, probeQ, sink,
sinkC, non_mut_target
]]),
name='non_mut_state')
# Compute proportion of probe which is active in each tube
result = nu.tube_analysis(tubes=[initial_state, mut_state, non_mut_state],
compute=['pairs', 'mfe'],
model=model)
initial_activation = parse_nupack_output(result, 'initial_state',
concentrations)
mut_activation = parse_nupack_output(result, 'mut_state', concentrations)
non_mut_activation = parse_nupack_output(result, 'non_mut_state',
concentrations)
return initial_activation, mut_activation, non_mut_activation
def test_stack_energy():
mod = nu.Model()
pairs = 'AT CG GC TA GT TG'.split()
pairs = 'AT CG GC TA'.split()
s = ''
for i in pairs:
for j in pairs:
seqs = [i[0] + j[0], j[1] + i[1]]
s += '%d ' % int(100*mod.loop_energy(seqs))
s += '\n'
return s
def basic(
ths: str,
rbs: str = "AGAGGAGA",
const_rna: Optional[Collection[str]] = None,
max_size: int = 3,
n_samples: int = 100,
celsius: float = 37.0,
filename: str = "output.txt",
nupack_v: int = 4,
) -> None:
"""Basic test against miRNAs from the H**o sapiens miRBase which saves results as text. You can also set file ending to .csv for a spreadsheet."""
if nupack_v == 4:
model = nupack.Model(celsius=celsius)
elif nupack_v == 3:
model = nupack.Model(ensemble="some-nupack3",
material="rna95-nupack3",
celsius=celsius)
else:
raise ValueError(
f"only NUPACK versions supported allowed, not 'nupack_v={nupack_v}'"
)
fasta = FASTA.fromspecies("H**o sapiens")
t = autoconfig(
ths=ths,
rbs=rbs,
triggers=fasta.seqs,
set_size=1,
names=fasta.IDs,
const_rna=const_rna,
model=model,
)
results = t.run(max_size, n_samples=n_samples)
with open(filename, "w", encoding="utf-8-sig") as f:
if filename.endswith(".csv"):
f.write(results.to_csv())
else:
f.write(results.prettify())
def autoconfig(
ths: str,
rbs: str,
triggers: Collection[str],
set_size: int = 1,
const_rna: Optional[Iterable[str]] = None,
names: Optional[Collection[str]] = None,
model: nupack.Model = nupack.Model(),
) -> ToeholdTest:
"""
Quick configuration of ToeholdTests, assuming every RNA has the same concentration.
Args:
ths (str): the toehold switch
rbs (str): the ribosome binding site whose position in the ths will be autodetected
triggers (list): an array-like of the triggers to test the ths against
set_size (int, optional): how many triggers to test against the ths at a time. Defaults to 1.
const_rna (list, optional): a list of any constant RNAs. Defaults to [].
names (list, optional): the names of each trigger. Defaults to [].
model (nupack.Model, optional): the physical model to use. Defaults to nupack.Model().
Returns:
a fully configured ToeholdTest instance
"""
rbs_location = find_rbs(ths, rbs)
trigger_sets = combs(triggers, set_size)
conc_sets = np.full(trigger_sets.shape, 1, dtype=np.float64)
t = ToeholdTest(
ths={ths: 1},
rbs_location=rbs_location,
trigger_sets=trigger_sets,
conc_sets=conc_sets,
const_rna={rna: 1
for rna in const_rna},
model=model,
)
if len(names) > 0:
assert len(names) == len(
triggers
), "each name must match to a corresponding trigger. Set to [] otherwise."
names = combs(names, set_size)
t.names = names
return t
import numpy as np
import pandas as pd
import nupack
# configure nupack model
NUPACK_MODEL = nupack.Model(
material=snakemake.config.get('nupack_material', 'dna'),
ensemble=snakemake.config.get('nupack_ensemble', 'stacking'),
celsius=snakemake.config.get('nupack_celsius', 74.5),
sodium=snakemake.config.get('nupack_sodium', 0.39),
magnesium=snakemake.config.get('nupack_magnesium', 0.0),
)
def main():
# load probe data
df = pd.read_csv(snakemake.input[0], sep='\t', header=None)
df.columns = [
'chrom',
'start',
'stop',
'seq',
'Tm',
'on_target_score',
'off_target_score',
'repeat'
]
# compute prob value for each sequence
df['prob'] = df['seq'].apply(prob)
def test_loop_energy():
em = nu.Model(ensemble='some-nupack3', material='rna95-nupack3')
assert abs(em.beta - 1.6225317071110157) < 1e-8
assert em.hairpin_energy('AAAAT') == 4.1
def test_structure_energy():
model = nu.Model(ensemble='some-nupack3', material='rna95-nupack3')
assert model.structure_energy('CCCCTTTGGGG', '((((...))))') == -4.6
assert model.structure_energy('GGAAACC', '.(...).') == 3.3
assert abs(nu.Model(ensemble='stacking', material='rna95-nupack3').structure_energy('GGAAACC', '.(...).') - 2.7239996321033035) < 1e-8