def find_discriminating_primers(cls,
alignment,
plen=25,
max_mismatches=3,
min_match_mismatches=3,
first_match_mismatches=3,
first_may_match=0,
AT_first=True,
outgroup=1,
reverse=False):
primers = []
if reverse: alignment = alignment.reverse_complement()
ali_len = alignment.get_alignment_length()
if isinstance(plen, (tuple, list)):
min_len = plen[0]
max_len = plen[1]
else:
min_len = max_len = plen
for s in alignment.cols:
if s < min_len - 1: continue
mlen = min_len
mis = 0
mmis = min_match_mismatches
pali = alignment[:, max(0, s - max_len):s]
if AT_first and pali[0, -1].upper() not in 'AT': continue
good = True
cur_len = 0
first_matched = 0
for i in pali.rcols:
cur_len += 1
col = pali[:, i]
group = col[:-outgroup]
if cur_len <= first_match_mismatches + first_matched:
if cur_len <= first_may_match:
if not issingleletter(group):
good = False
break
elif not cls._match_mismatch(col, outgroup):
first_matched += 1
elif not cls._match_mismatch(col, outgroup):
good = False
break
continue
if '-' in group:
if issingleletter(group, '-'): mlen += 1
else:
good = False
break
if cls._match_mismatch(col, outgroup): mmis -= 1
if not issingleletter(group): mis += 1
if cur_len >= mlen and mmis <= 0 and mis <= max_mismatches:
break
good &= cur_len >= mlen and mmis <= 0 and mis <= max_mismatches
if not good: continue
pos = ali_len - s if reverse else s + 1
primers.append(
(pos, pali[:-outgroup,
pali.get_alignment_length() - cur_len:].trim()))
return primers
def _find_primers(self, alignment,
plen = 25,
max_mismatches = 3,
min_match_mismatches = 3,
first_match_mismatches = 3,
first_may_match = 0,
AT_first = True,
outgroup=1):
primers = []
if isinstance(plen, (tuple, list)):
min_len = plen[0]
max_len = plen[1]
else: min_len = max_len = plen
for s in alignment.cols:
if s < min_len-1: continue
mlen = min_len
mis = 0
mmis = min_match_mismatches
pali = alignment[:,max(0,s-max_len):s]
if AT_first and pali[0,-1].upper() not in 'AT': continue
good = True
cur_len = 0
first_matched = 0
for i in pali.rcols:
cur_len += 1
col = pali[:,i]
group = col[:-outgroup]
if cur_len <= first_match_mismatches+first_matched:
if cur_len <= first_may_match:
if not issingleletter(group):
good = False
break
elif not self._match_mismatch(col, outgroup):
first_matched += 1
elif not self._match_mismatch(col, outgroup):
good = False
break
continue
if '-' in group:
if issingleletter(group, '-'): mlen += 1
else:
good = False
break
if self._match_mismatch(col, outgroup): mmis -= 1
if not issingleletter(group): mis += 1
if cur_len >= mlen and mmis <= 0 and mis <= max_mismatches: break
good &= cur_len >= mlen and mmis <= 0 and mis <= max_mismatches
if not good: continue
primers.append((s+1, pali[:-outgroup,pali.get_alignment_length()-cur_len:].trim()))
print '\nposition: %d\n%s' % primers[-1]#test
return primers
def trim(self, gap = '-'):
'''Trim alignment termina containing gaps in columns
and columns containing only gaps'''
#trim margins
lmargin = 0
for i in self.cols:
lmargin = i
if gap not in self[:,i]: break
if lmargin == self.get_alignment_length()-1:
return AlignmentExt([])
rmargin = -1
for i in self.rcols:
rmargin = i+1
if gap not in self[:,i]: break
ali = self[:,lmargin:rmargin]
#remove gaps-only
cols = set(ali.cols)
for i in ali.cols:
if issingleletter(ali[:,i], gap):
cols.remove(i)
cols = sorted(cols)
slices = []
s = 0
for i in xrange(len(cols)-1):
if cols[i+1]-cols[i] > 1:
slices.append(slice(s, cols[i]+1))
s = cols[i+1]
slices.append(slice(s, cols[-1]+1))
final = ali[:,slices[0]]
for s in slices[1:]: final += ali[:,s]
return final
def trim(self, gap='-'):
'''Trim alignment termina containing gaps in columns
and columns containing only gaps'''
#trim margins
lmargin = 0
for i in self.cols:
lmargin = i
if gap not in self[:, i]: break
if lmargin == self.get_alignment_length() - 1:
return AlignmentExt([])
rmargin = -1
for i in self.rcols:
rmargin = i + 1
if gap not in self[:, i]: break
ali = self[:, lmargin:rmargin]
#remove gaps-only
cols = set(ali.cols)
for i in ali.cols:
if issingleletter(ali[:, i], gap):
cols.remove(i)
cols = sorted(cols)
slices = []
s = 0
for i in xrange(len(cols) - 1):
if cols[i + 1] - cols[i] > 1:
slices.append(slice(s, cols[i] + 1))
s = cols[i + 1]
slices.append(slice(s, cols[-1] + 1))
final = ali[:, slices[0]]
for s in slices[1:]:
final += ali[:, s]
return final
def find_specific_primers(alignment,
plen=25,
max_mismatches=3,
min_first_matches=3,
AT_first=True,
reverse=False):
primers = []
if reverse: alignment = alignment.reverse_complement()
ali_len = alignment.get_alignment_length()
if isinstance(plen, (tuple, list)):
min_len = plen[0]
max_len = plen[1]
else:
min_len = max_len = plen
for s in alignment.cols:
if s < min_len - 1: continue
mlen = min_len
mis = 0
pali = alignment[:, max(0, s - max_len):s]
if AT_first and pali[0, -1].upper() not in 'AT': continue
good = True
cur_len = 0
for i in pali.rcols:
cur_len += 1
col = pali[:, i]
if cur_len <= min_first_matches:
if not issingleletter(col):
good = False
break
continue
if '-' in col:
if issingleletter(col, '-'): mlen += 1
else:
good = False
break
if not issingleletter(col): mis += 1
if mis > max_mismatches:
if cur_len > mlen: cur_len -= 1
else: good = False
break
if not good: continue
pali_len = pali.get_alignment_length()
if not issingleletter(pali[:, pali_len - cur_len]): cur_len -= 1
primers.append((ali_len - s if reverse else s + 1,
pali[:, pali_len - cur_len:].trim()))
return primers
def _match_mismatch(self, s, outgroup=1):
if not s: return False
return (issingleletter(s[:-outgroup])
and not s[0] in s[-outgroup:])
def _match_mismatch(s, outgroup=1):
if not s: return False
return (issingleletter(s[:-outgroup]) and not s[0] in s[-outgroup:])