def penalize_solution(best_solution: SSMSolution, config: SSMConfig,
fw_opt_temp, rv_opt_temp, flanks: SSMFlankingSequences):
"""
Computes primer dimer penalty. The penalty consists of following penalties:
- hairpin penalty
- h**o dimer penalty
- hetero dimer penalty (this is applied for each combinations of primer pairs with the rest)
Each component is computed with help of primer3 library. It is weighted squared error by corresponding weights
from config. Than all errors are added together and we return square error of this errors.
:param best_solution:
:param config:
:param fw_opt_temp:
:param rv_opt_temp:
:return:
"""
for pair in best_solution.result:
fw_sequence = pair.fw_primer.normal_order_sequence
rw_sequence = pair.rw_primer.normal_order_sequence
temp_cfg = config.temperature_config
fw_hairpin_tm = primer3.calcHairpinTm(fw_sequence, temp_cfg.k,
temp_cfg.mg, temp_cfg.dntp)
rw_hairpin_tm = primer3.calcHairpinTm(rw_sequence, temp_cfg.k,
temp_cfg.mg, temp_cfg.dntp)
fw_homodimer_tm = primer3.calcHomodimerTm(fw_sequence, temp_cfg.k,
temp_cfg.mg, temp_cfg.dntp)
rw_homodimer_tm = primer3.calcHomodimerTm(rw_sequence, temp_cfg.k,
temp_cfg.mg, temp_cfg.dntp)
fw_hairpin_err = (best_solution.forward_temp - fw_hairpin_tm)**2
rw_hairpin_err = (best_solution.reverse_temp - rw_hairpin_tm)**2
fw_homodimer_err = (fw_opt_temp - fw_homodimer_tm)**2
rw_homodimer_err = (rv_opt_temp - rw_homodimer_tm)**2
heterodimer_err = compute_heterodimer_err(pair, config, flanks)
penalty = math.sqrt(
config.hairpin_temperature_weight * fw_hairpin_err +
config.hairpin_temperature_weight * rw_hairpin_err +
config.primer_dimer_temperature_weight * fw_homodimer_err +
config.primer_dimer_temperature_weight * rw_homodimer_err +
config.primer_dimer_temperature_weight * heterodimer_err)
pair.non_optimality += penalty
def primer3_filter_withRprimer(sequence, rprimer, mintm=37, maxhtm=35, dtm=10):
primer3ft = False
tm = primer3.calcTm(sequence)
fseq = rprimer + sequence
htmF = primer3.calcHairpinTm(fseq)
rseq = rprimer + revcom(sequence)
htmR = primer3.calcHairpinTm(rseq)
if tm < mintm:
primer3ft = True
if htmF > maxhtm:
primer3ft = True
if (tm-htmF) < dtm:
primer3ft = True
if htmR > maxhtm:
primer3ft = True
if (tm-htmR) < dtm:
primer3ft = True
# print(sequence, tm, htm, dtm)
return primer3ft
def is_good_primer(primer):
# ref1. http://www.premierbiosoft.com/tech_notes/PCR_Primer_Design.html
seq = ''.join([i[1] for i in primer])
if re.search(poly, seq) is not None:
return False, 0, 'Poly(NNNNN) structure found'
if re.search(tandem, seq) is not None:
return False, 0, 'Tandom(NN*5) exist'
# no more 3 ambiguous base
if len(re.findall(ambiguous_base, seq)) >= ambiguous_base_n:
return False, 0, 'More than 3 ambiguous base'
# primer3.setGlobals seems have no effect on calcTm, so I have to replace all
# ambiguous base to A to get an approximate value. Othervise calcTm() will
# generate -99999 if there is ambiguous base.
pure_seq = re.sub(ambiguous_base, 'A', seq)
tm = primer3.calcTm(pure_seq)
hairpin_tm = primer3.calcHairpinTm(pure_seq)
homodimer_tm = primer3.calcHomodimerTm(pure_seq)
if max(tm, hairpin_tm, homodimer_tm) != tm:
return False, 0, 'Hairpin or homodimer found'
return True, tm, 'Ok'
def primer3_cal(sequence, mintm=37, maxhtm=37, dtm=10):
primer3ft = True
tm = primer3.calcTm(sequence)
htm = primer3.calcHairpinTm(sequence)
if tm < mintm:
primer3ft = False
if htm > maxhtm:
primer3ft = False
if (tm-htm) > dtm:
primer3ft = False
return (sequence, primer3ft)
def primer3filter(seq, min_TM=37, max_HTM=35, min_diff_TM=10):
# Calculate
TM = primer3.calcTm(seq)
HTM = primer3.calcHairpinTm(seq)
# If melting temperature is too low, filter out
if TM < min_TM:
return "melting temp too low"
# If hairpin melting temperature is too high, filter out
elif HTM > max_HTM:
return "hairpin melting temp too high"
# If melting temperature and hairpin melting temperature
# are too close together, filter out
elif (TM - HTM) < min_diff_TM:
return "difference between melting temp and hairpin melting temp too small"
# If sequence will make a good probe, return false (do not filter)
else:
return False
def primer3_filter(sequence, mintm=37, maxhtm=35, dtm=10):
primer3ft = False
tm = primer3.calcTm(sequence)
htm = primer3.calcHairpinTm(sequence)
if tm < mintm:
primer3ft = True
if htm > maxhtm:
primer3ft = True
if (tm-htm) < dtm:
primer3ft = True
# print(sequence, tm, htm, dtm)
return primer3ft
def primer3_filter(line, min_TM, max_HTM, min_diff_TM):
# Get sequence from line passed to function
sequence = line.split('\t')[9]
# Calculate
TM = primer3.calcTm(sequence)
HTM = primer3.calcHairpinTm(sequence)
# If melting temperature is too low, filter out
if TM < min_TM:
return True
# If hairpin melting temperature is too high, filter out
elif HTM > max_HTM:
return True
# If melting temperature and hairpin melting temperature
# are too close together, filter out
elif (TM - HTM) < min_diff_TM:
return True
# If sequence will make a good probe, return false (do not filter)
else:
return False
