def DSeqVH(fwd: str,
rev: str = None,
overhang: int = None,
alphabet: int = AlphaEnum.DNA) -> VirtualHelix:
'''Helper function for creating :class:`VirtualHelix` in the style of
the :class:`DSeq` with strings
'''
dseq: DSeq = DSeq(fwd, rev, overhang, alphabet)
overhang: int = dseq.overhang
if overhang > 0:
fwd_idx_offsets = [overhang]
rev_idx_offsets = [0]
else:
fwd_idx_offsets = [0]
rev_idx_offsets = [overhang]
oligo_fwd = Oligo(fwd)
if rev is None:
rev = reverseComplement(fwd)
oligo_rev = Oligo(rev)
return VirtualHelix([oligo_fwd.strand5p], fwd_idx_offsets,
[oligo_rev.strand5p], rev_idx_offsets)
def find_tms(cls,
sequences,
sodium=0.05,
magnesium=0.0,
temperature=25,
concentration=0.00000025,
**kwargs):
cls.load()
def flatten(iterable, remove_none=False, add_equal=False):
"""Make a flat list out of a list of lists"""
# Will remove None
if remove_none:
if add_equal:
return [
'{}={}'.format(par, val) if (val != None) else par
for (par, val) in iterable
]
else:
return [
item for sublist in iterable for item in sublist
if item != None
]
else:
if add_equal:
return [
'{}={}'.format(par, val) for (par, val) in iterable
]
else:
return [item for sublist in iterable for item in sublist]
# RNAplex in 'probe mode' only calculates the reverse-complement Tm, and cannot
# calculate the hairpin, homodimer, or heterodimer Tms.
options = OrderedDict([
('--paramFile', cls.parameters),
('--probe-mode', None),
('--probe-concentration', Oligo.float_to_str(concentration)),
('--na-concentration', Oligo.float_to_str(sodium)),
('--mg-concentration', Oligo.float_to_str(magnesium)),
('--tris-concentration', Oligo.float_to_str(0.0)),
('--k-concentration', Oligo.float_to_str(0.0)),
('--temp', temperature),
])
flat_options = flatten(options.items(),
remove_none=True,
add_equal=True)
command_list = ['RNAplex'] + list(map(str, flat_options))
command_str = ' '.join(command_list)
cls.logger.info('command: {!r}'.format(command_str))
cp = subprocess.run(command_list,
input=bytes('\n'.join(sequences), 'utf-8'),
shell=False,
stdout=subprocess.PIPE,
stderr=subprocess.DEVNULL)
# The output typically looks like this:
# Probe mode
# Concentration K:0.000 TNP:0.000 Mg:0.000 Na:0.050 probe:0.000
#
# sequence DDSL98 DDSL04 DRSU95 RRXI98 CURRENT
# AGGCTTTAGGGCTATAGGAA 51.78 50.76 50.74 62.06 52.13
# CGAATTTAGAGCCTATAAT 43.60 42.90 39.52 48.07 43.46
# GGCTATGAGATAGCTAA 43.14 42.68 41.24 52.81 43.44
out_lines = cp.stdout.decode().splitlines()
data_found = False
tm_list = []
for line in out_lines:
line = line.rstrip()
if not data_found:
m = regex.search(
r'^\s+sequence\s+DDSL98\s+DDSL04\s+DRSU95\s+RRXI98\s+CURRENT',
line)
if m:
data_found = True
else:
m = regex.match(r'^\s*(\S+)(?:\s+(\S+)){5}$', line)
if m:
seq = m.captures(1)[0]
tm = float(m.captures(2)[-1]) # Use the 'CURRENT' column
tm_list.append(tm)
return tm_list
def breakStrand(self, dir_idx: int, strand: Strand,
idx: int) -> Tuple[Oligo, Oligo, Oligo]:
'''Break a Strand in two and create two new Oligos to assign the all of
the strands in the pre-existing Oligo to
Args:
dir_idx: is this on the forward [0] or reverse [1] direction of the
:class:`VirtualHelix`. Also use :enum:`VHDirEnum` to get these idxs
strand: :class:`Strand` object to break
idx: index to break the strand at in terms of it's sequence
Returns:
Two new :class:`Oligo` objects of form::
Oligo_5p, Oligo_3p
'''
VHDirEnum.check(dir_idx)
strand_array = self.strand_arrays[dir_idx]
vh_strands: List[Strand] = strand_array.strands
if strand not in vh_strands:
dir_name: str = VHDirEnum(dir_idx).name
err: str = "Strand {} not in the {} StrandArray of the VirtualHelix"
raise ValueError(err.format(strand, dir_name))
idx_offsets: List[int] = strand_array.idx_offsets
seq: str = strand.seq
oligo_old: Oligo = strand.oligo
# 1. Do the 5' portion of the break
oligo_break5p: Oligo = Oligo(seq[0:idx])
strand_break5p: Strand = oligo_break5p.strand5p
neighbor_5p: Strand = strand.strand5p
if neighbor_5p is not None: # update existing neighbor oligos
strand_break5p.strand5p = neighbor_5p
neighbor_5p.strand3p = strand_break5p
for seg in neighbor_5p.gen5p():
seg.oligo = oligo_break5p
# 2. Do the 3' portion of the break
oligo_break3p: Oligo = Oligo(seq[idx:])
strand_break3p: Strand = oligo_break3p.strand5p
neighbor_3p: Strand = strand.strand3p
if neighbor_3p is not None: # update existing neighbor oligos
strand_break3p.strand3p = neighbor_3p
neighbor_3p.strand5p = strand_break3p
for seg in neighbor_3p.gen3p():
seg.oligo = oligo_break3p
# 3. Update the strands
list_idx: int = vh_strands.index(strand)
offset_5p: int = idx_offsets[list_idx]
list_idx_plus_1: int = list_idx + 1
vh_strands.insert(list_idx_plus_1, strand_break3p)
vh_strands.insert(list_idx_plus_1, strand_break5p)
idx_offsets.insert(list_idx_plus_1, offset_5p + len(strand_break5p))
vh_strands.pop(list_idx) # pop out the original strand
return oligo_break5p, oligo_break3p
def str2Oligo(x: str) -> Tuple[bool, Oligo, Strand]:
if isinstance(x, Strand):
return False, x.oligo, x
else:
oligo = Oligo(x)
return True, oligo, oligo.strand5p
else:
fwd_idx_offsets = [0]
rev_idx_offsets = [overhang]
oligo_fwd = Oligo(fwd)
if rev is None:
rev = reverseComplement(fwd)
oligo_rev = Oligo(rev)
return VirtualHelix([oligo_fwd.strand5p], fwd_idx_offsets,
[oligo_rev.strand5p], rev_idx_offsets)
# end def
if __name__ == '__main__':
fwd = 'GGTCTCGAATTCAAA'
oligo_fwd = Oligo(fwd)
rev = 'TTTGAATTCGAGACC'
oligo_rev = Oligo(rev)
BsaI_vh = VirtualHelix([oligo_fwd.strand5p], [0], [oligo_rev.strand5p],
[0])
print("1.\n%s" % BsaI_vh)
BsaI_vh = DSeqVH(fwd, rev, 0)
print("2.\n%s" % BsaI_vh)
print(BsaI_vh.fwd_strands)
BsaI_vh = DSeqVH(fwd)
print("3.\n%s" % BsaI_vh)
print("Da Oligos", BsaI_vh.oligos())
strand0 = BsaI_vh.fwd_strands[0]
print(strand0.oligo)
broken_oligos = BsaI_vh.breakStrand(dir_idx=0, strand=strand0, idx=4)
def nnn_unafold(cls,
folder,
seq1,
seq2=None,
sodium=0.05,
magnesium=0.0,
temperature=25,
concentration=0.00000025):
"""
Calculate deltaG, deltaH, deltaS, Tm.
Faster than running 'UNAFold.pl'.
Accepts 1 or 2 input sequences. Automatically runs either:
* Hairpin (1 input sequence: A=seq1, UNAFold run on A)
* Homodimer (2 identical input sequences: A=seq1=seq2, UNAFold run on A & A)
* Heterodimer (2 input sequences: A=seq1 B=seq2, UNAFold run on A & B)
Writes '*.det' file to temp 'folder'
Returns four lists:
([deltaG, ...], [deltaH, ...], [deltaS, ...], [Tm, ...])
"""
# Create sequence files for input
with open(os.path.join(folder, 'A.seq'), 'w') as flo:
print(seq1, file=flo)
if seq2:
if (seq2 != seq1):
with open(os.path.join(folder, 'B.seq'), 'w') as flo:
print(seq2, file=flo)
# 1 sequence
# @command = ('hybrid-ss', @rules, @rules2, @rules3, '--tracebacks'=> $max);
# Only used when --model=PG
#hybrid-ss --NA DNA --tmin 25 --tmax 25 --sodium=0.05 --magnesium=0.0 --suffix DAT A.seq # <-- not actually tested yet
#output_basename=None # add to function arguments
#if output_basename:
# outfile = os.path.join(folder, output_basename)
#else:
# outfile = os.devnull
parameters_1 = [
'--NA=DNA',
'--tmin=' + str(temperature),
'--tmax=' + str(temperature),
'--tinc=1',
'--sodium=' + Oligo.float_to_str(sodium),
'--magnesium=' + Oligo.float_to_str(magnesium),
'--maxloop=' + str(30),
'--mfold=' + ','.join(map(str, [5, -1, 100])),
]
if (seq2 == None):
prefix = 'A'
# 1 sequence
# Used by default, or when --model=EM
# Creates files: A.ann, A.ct, A.dG, A.plot, A.run
# Command: hybrid-ss-min --NA=DNA --tmin=25 --tmax=25 --tinc=1 --sodium=0.05 --magnesium=0.0 --maxloop=30 --mfold=5,-1,100 A.seq
command_list = ['hybrid-ss-min'] + parameters_1 + ['A.seq']
elif (seq1 == seq2):
prefix = 'A-A'
command_list = ['hybrid-min'] + parameters_1 + ['A.seq', 'A.seq']
else:
prefix = 'A-B'
command_list = ['hybrid-min'] + parameters_1 + ['A.seq', 'B.seq']
try:
with open(os.devnull, 'w+') as flo: # Prevent printing to STDOUT
cp = subprocess.run(command_list,
shell=False,
check=True,
cwd=folder,
stdout=flo,
stderr=subprocess.STDOUT)
except subprocess.CalledProcessError:
# Some sequences, such as 'GAGAAGGAGAAGGAGAAG' paired with itself, will cause a segmentation fault
return [math.inf], [math.inf], [math.nan], [math.nan]
# If model=EM
# Apparently, this is not needed
# Creates file: A.h-num
# Command: h-num.pl A
# with open(outfile, 'w+') as flo:
# command_list = ['h-num.pl', 'A']
# cp = subprocess.run(command_list, shell=False, check=True, cwd=folder, stdout=flo, stderr=subprocess.STDOUT)
# Apparently, this is not needed
# Creates file: A.ss-count
# Command: ss-count.pl A.ct > A.ss-count
# with open('A.ss-count', 'w') as flo:
# command_list = ['ss-count.pl', 'A.ct']
# cp = subprocess.run(command_list, shell=False, check=True, cwd=folder, stdout=flo, stderr=None)
# This calculates delta-H
# if there is no --suffix option used '--suffix=DHD' overrides '--NA=DNA --sodium=0.05 --magnesium=0.0 --temperature=25'
# Command: ct-energy --suffix=DHD A.ct > A.deltaH
command_list = ['ct-energy', '--suffix=DHD', prefix + '.ct']
cp = subprocess.run(command_list,
shell=False,
check=True,
cwd=folder,
stdout=subprocess.PIPE,
stderr=None)
deltaH_list = []
for line in cp.stdout.decode().splitlines():
deltaH_list.append(float(line))
# This gives more precise delta-G calculations
# Command: ct-energy --NA=DNA --sodium=0.05 --magnesium=0.0 --temperature=25 A.ct > A.deltaG
parameters_2 = [
'--NA=DNA',
'--temperature=' + str(temperature),
'--sodium=' + Oligo.float_to_str(sodium),
'--magnesium=' + Oligo.float_to_str(magnesium),
]
command_list = ['ct-energy'] + parameters_2 + [prefix + '.ct']
cp = subprocess.run(command_list,
shell=False,
check=True,
cwd=folder,
stdout=subprocess.PIPE,
stderr=None)
deltaG_list = []
for line in cp.stdout.decode().splitlines():
deltaG_list.append(float(line))
# This gives the complicated information within the A.det file...
#ct-energy --NA=DNA --sodium=0.05 --magnesium=0.0 --temperature=25 --verbose A.ct | ct-energy-det.pl --mode text
number_structures = len(deltaH_list)
homodimer_list = cls.parse_ct_file(os.path.join(
folder, prefix + '.ct'
)) # Check each pair to determine if it's a h**o- or heterodimer
#deltaH_list = [-32.4, -49.8]
#deltaG_list = [-3.43067, -2.56959]
deltaS_list = []
Tm_list = []
factor_list = []
#temperature = 25 # argument
#concentration = 0.00000025 # argument
R = 0.0019872 # Constant
try:
for i in range(number_structures):
#if (len(homodimer_list) > i):
# homodimer = homodimer_list[i]
#else:
# homodimer = None
homodimer = homodimer_list[i]
deltaH = deltaH_list[i]
deltaG = deltaG_list[i]
deltaS = 1000.0 * (deltaH - deltaG) / (273.15 + temperature)
if (homodimer != None):
if (homodimer == True):
factor = 1
else:
factor = 4
Tm = 1000.0 * deltaH / (
deltaS + 1000.0 * R * math.log(concentration / factor)
) - 273.15 # Natural log (base e)
else:
Tm = 1000.0 * deltaH / deltaS - 273.15
factor = None
deltaS_list.append(deltaS)
Tm_list.append(Tm)
factor_list.append(factor)
except IndexError:
# Problematic sequences, such as 'TTCTCCACTTCCATCACC' will cause an error
# So we artificially give them unreasonable results so they will be discarded
for i in range(number_structures):
deltaS_list.append(math.nan)
Tm_list.append(math.nan)
factor_list.append(None)
# Write the det file
with open(os.path.join(folder, prefix + '.det'), 'w') as flo:
for i in range(number_structures):
print(
'Structure {}: dG = {} dH = {} dS = {} Tm = {}'.format(
i + 1, deltaG_list[i], deltaH_list[i], deltaS_list[i],
Tm_list[i]),
file=flo)
# Create Probability dot plot
#my @command = ('hybrid-plot-ng', '--temperature' => $temp)
#system(@command)
# Create Energy dot plot
#my @command = ('boxplot_ng', '-d', -c => 4);
#system(@command)
# Create structure plots
#system($sirgraph, @flags, -ss => "${prefix}_$fold")
#system($sirgraph, @flags, -p => "${prefix}_$fold")
#system($sirgraph, @flags, $img, "${prefix}_$fold")
#system('ps2pdfwr', "${prefix}_$fold.ps")
# ????
#system('ct2rnaml', $prefix)
# These are all of the float data type
return deltaG_list, deltaH_list, deltaS_list, Tm_list
def calculate_simple(cls,
folder_p,
seq1,
seq2=None,
sodium=0.05,
magnesium=0.0,
temperature=25,
concentration=0.00000025,
output_basename=None):
"""
Writes files to temp folder 'folder_p'
If 1 input sequence, then UNAFold is run on seq1 only
If 2 input sequences, then UNAFold run on seq1+seq2
"""
# Make temporary folder
folder = os.path.join(folder_p, 'oligos')
os.makedirs(folder, exist_ok=True)
# Create sequence files for input
with open(os.path.join(folder, 'A.seq'), 'w') as flo:
print(seq1, file=flo)
if seq2:
with open(os.path.join(folder, 'B.seq'), 'w') as flo:
print(seq2, file=flo)
# Default concetrations: 0.25 uM = 0.00025 mM = 0.00000025 M
basic_command_list = [
'UNAFold.pl', '--NA=DNA', '--temp=' + str(temperature),
'--sodium=' + Oligo.float_to_str(sodium),
'--magnesium=' + Oligo.float_to_str(magnesium),
'--Ct=' + Oligo.float_to_str(concentration), '--max=100'
]
if output_basename:
outfile = os.path.join(folder, output_basename)
else:
outfile = os.devnull
with open(outfile, 'w+') as flo:
if seq2:
# Do heterodimer
command_list = basic_command_list + ['A.seq', 'B.seq']
cp = subprocess.run(command_list,
shell=False,
check=True,
cwd=folder,
stdout=flo,
stderr=subprocess.STDOUT)
else:
# Do hairpins
try:
command_list = basic_command_list + ['A.seq']
cp = subprocess.run(command_list,
shell=False,
check=True,
cwd=folder,
stdout=flo,
stderr=subprocess.STDOUT)
except subprocess.CalledProcessError: # This error happens sometimes, i.e. when A.seq contains 'TTCTCCACTTCCATCACC'
# UNAFold.pl crashes, so we just write a crappy 'A.det' file
print("UNAFold.pl CRASH")
with open(os.path.join(folder, 'A.det'), 'w') as crash_flo:
print(
'Structure 1: dG = -999.0 dH = -999.0 dS = -999.0 Tm = 99.0',
file=crash_flo)
# Make the objects
if seq2:
return cls.make_objects(os.path.join(folder,
'A-B.det'), sodium, magnesium,
temperature, concentration, seq1, seq2)
else:
return cls.make_objects(os.path.join(folder,
'A.det'), sodium, magnesium,
temperature, concentration, seq1)
def calculate_full(cls,
folder_p,
seq1,
seq2,
sodium=0.05,
magnesium=0.0,
temperature=25,
concentration=0.00000025,
output_basename=None):
# Make temporary folder
folder = os.path.join(folder_p, 'oligos')
os.makedirs(folder, exist_ok=True)
# Create sequence files for input
with open(os.path.join(folder, 'A.seq'), 'w') as flo:
print(seq1, file=flo)
with open(os.path.join(folder, 'B.seq'), 'w') as flo:
print(seq2, file=flo)
with open(os.path.join(folder, 'rcA.seq'), 'w') as flo:
print(rc(seq1), file=flo)
with open(os.path.join(folder, 'rcB.seq'), 'w') as flo:
print(rc(seq2), file=flo)
# Default concetrations: 0.25 uM = 0.00025 mM = 0.00000025 M
basic_command_list = [
'UNAFold.pl', '--NA=DNA', '--temp=' + str(temperature),
'--sodium=' + Oligo.float_to_str(sodium),
'--magnesium=' + Oligo.float_to_str(magnesium),
'--Ct=' + Oligo.float_to_str(concentration), '--max=100'
]
if output_basename:
outfile = os.path.join(folder, output_basename)
else:
outfile = os.devnull
with open(outfile, 'w+') as flo:
# Do hairpins
command_list = basic_command_list + ['A.seq']
cp = subprocess.run(command_list,
shell=False,
check=True,
cwd=folder,
stdout=flo,
stderr=subprocess.STDOUT)
command_list = basic_command_list + ['B.seq']
cp = subprocess.run(command_list,
shell=False,
check=True,
cwd=folder,
stdout=flo,
stderr=subprocess.STDOUT)
# Do homodimers
command_list = basic_command_list + ['A.seq', 'A.seq']
cp = subprocess.run(command_list,
shell=False,
check=True,
cwd=folder,
stdout=flo,
stderr=subprocess.STDOUT)
command_list = basic_command_list + ['B.seq', 'B.seq']
cp = subprocess.run(command_list,
shell=False,
check=True,
cwd=folder,
stdout=flo,
stderr=subprocess.STDOUT)
# Do heterodimer
command_list = basic_command_list + ['A.seq', 'B.seq']
cp = subprocess.run(command_list,
shell=False,
check=True,
cwd=folder,
stdout=flo,
stderr=subprocess.STDOUT)
# Do reverse complements
command_list = basic_command_list + ['A.seq', 'rcA.seq']
cp = subprocess.run(command_list,
shell=False,
check=True,
cwd=folder,
stdout=flo,
stderr=subprocess.STDOUT)
command_list = basic_command_list + ['B.seq', 'rcB.seq']
cp = subprocess.run(command_list,
shell=False,
check=True,
cwd=folder,
stdout=flo,
stderr=subprocess.STDOUT)
# Make the objects
#for det_filename in ['A.det', 'B.det', 'A-A.det', 'B-B.det', 'A-B.det']:
# pass
a = cls.make_objects(os.path.join(folder, 'A.det'), sodium, magnesium,
temperature, concentration, seq1)
b = cls.make_objects(os.path.join(folder, 'B.det'), sodium, magnesium,
temperature, concentration, seq2)
aa = cls.make_objects(os.path.join(folder,
'A-A.det'), sodium, magnesium,
temperature, concentration, seq1, seq1)
bb = cls.make_objects(os.path.join(folder,
'B-B.det'), sodium, magnesium,
temperature, concentration, seq2, seq2)
ab = cls.make_objects(os.path.join(folder,
'A-B.det'), sodium, magnesium,
temperature, concentration, seq1, seq2)
ra = cls.make_objects(os.path.join(folder, 'A-rcA.det'), sodium,
magnesium, temperature, concentration, seq1,
rc(seq1))
rb = cls.make_objects(os.path.join(folder, 'B-rcB.det'), sodium,
magnesium, temperature, concentration, seq2,
rc(seq2))
return a, b, aa, bb, ab, ra, rb