def pcr_sequences_for_amplicon(amplicon, padding_pos5=0, padding_pos3=0, include_snps=False):
"""
Returns the PCRSequence objects representing the specified amplicon.
If include_snps=True, each object will include an attribute 'snps', which will
include a list of SNP dictionaries (like the SNPDBCache object, only
SQLAlchemy-agnostic, and 'class' in place of 'class_')
:param padding_pos5: The amount of padding to prefix the amplicon. Max MAX_CACHE_PADDING.
:param padding_pos3: The amount of padding to suffix the amplicon. Max MAX_CACHE_PADDING.
:param include_snps: Whether to include the snps attribute on each amplicon.
"""
if padding_pos5 < 0 or padding_pos5 > MAX_CACHE_PADDING:
raise ValueError, "Illegal padding value: %s" % padding_pos5
if padding_pos3 < 0 or padding_pos3 > MAX_CACHE_PADDING:
raise ValueError, "Illegal padding value: %s" % padding_pos3
pseqs = []
for seq in amplicon.cached_sequences:
main = SimpleGenomeSequence(seq.chromosome, seq.start_pos, seq.end_pos, '+', seq.positive_amplicon)
# this could be of a different length than requested
padding_pos5_seq = seq.padding_pos5(padding_pos5, '+')
padding_pos3_seq = seq.padding_pos3(padding_pos3, '+')
prefix = SimpleGenomeSequence(seq.chromosome,
seq.start_pos-len(padding_pos5_seq),
seq.start_pos-1,
'+', padding_pos5_seq)
suffix = SimpleGenomeSequence(seq.chromosome,
seq.end_pos+1,
seq.end_pos+len(padding_pos3_seq),
'+', padding_pos3_seq)
pseq = PCRSequence(main, prefix, suffix)
if include_snps:
snps = seq.snps_in_range(padding_pos5=len(padding_pos5_seq), padding_pos3=len(padding_pos3_seq))
pseq.snps = [dict([(k, v) for k, v in snp.__dict__.items() if not k.startswith('_')]) for snp in snps]
for snp in pseq.snps:
snp['class'] = snp['class_']
pseqs.append(pseq)
return pseqs
def sequences_snps_for_assay(config, assay, seq_source, snp_source, left_padding=0, right_padding=0, cache=True):
sequences = []
if not assay:
return sequences
if assay.cached_sequences:
all_cached = True
for seq in assay.cached_sequences:
if not seq.cached(left_padding, right_padding):
all_cached = False
break
if all_cached:
for seq in assay.cached_sequences:
amplicon = SimpleGenomeSequence(seq.chromosome, seq.start_pos, seq.end_pos, "+", seq.positive_amplicon)
prefix = SimpleGenomeSequence(
seq.chromosome,
seq.start_pos - left_padding,
seq.start_pos - 1,
"+",
seq.padding_pos5(left_padding, "+"),
)
suffix = SimpleGenomeSequence(
seq.chromosome,
seq.end_pos + 1,
seq.end_pos + right_padding,
"+",
seq.padding_pos3(right_padding, "+"),
)
pseq = PCRSequence(amplicon, prefix, suffix)
# TODO unify SNP object
pseq.snps = [
dict([(k, v) for k, v in snp.__dict__.items() if not k.startswith("_")]) for snp in seq.snps
]
for snp in pseq.snps:
snp["class"] = snp["class_"]
sequences.append(pseq)
return sequences
if assay.assay_type == Assay.TYPE_PRIMER:
sequences = seq_source.sequences_for_primers(assay.primer_fwd, assay.primer_rev, left_padding, right_padding)
elif assay.assay_type == Assay.TYPE_LOCATION:
sequence = seq_source.sequence_around_loc(
assay.chromosome, assay.probe_pos, assay.amplicon_width, left_padding, right_padding
)
sequences.append(sequence)
elif assay.assay_type == Assay.TYPE_SNP:
snps = snp_source.snps_by_rsid(assay.snp_rsid)
# TODO: make SNP object so that access style is same as assay
for snp in snps:
if snp["refUCSC"] == "-": # deletion:
sequences.append(
seq_source.sequence_around_region(
snp["chrom"][3:],
snp["chromEnd"],
snp["chromEnd"],
assay.amplicon_width,
left_padding,
right_padding,
)
)
else:
sequences.append(
seq_source.sequence_around_region(
snp["chrom"][3:],
snp["chromStart"] + 1,
snp["chromEnd"],
assay.amplicon_width,
left_padding,
right_padding,
)
)
for seq in sequences:
seq.snps = snp_source.snps_in_range(seq.chromosome, seq.start, seq.end)
if cache:
# TODO: library method? -- given PCR object, SNP dict? or unify display objects
# to and from their DB representation?
assay.cached_sequences = []
for seq in sequences:
cached_seq = HG19AssayCache(
chromosome=seq.chromosome,
start_pos=seq.amplicon.start,
end_pos=seq.amplicon.end,
seq_padding_pos5=left_padding,
seq_padding_pos3=right_padding,
positive_sequence=seq.merged_positive_sequence.sequence,
)
cached_seq.amplicon_dg = dg_seq(config, cached_seq.positive_amplicon)
cached_seq.amplicon_tm = tm_probe(config, cached_seq.positive_amplicon)
for snp in seq.snps:
cached_seq.snps.append(
SNP131AssayCache(
bin=snp["bin"],
chrom=snp["chrom"],
chromStart=snp["chromStart"],
chromEnd=snp["chromEnd"],
name=snp["name"],
score=snp["score"],
strand=snp["strand"],
refNCBI=snp["refNCBI"],
refUCSC=snp["refUCSC"],
observed=snp["observed"],
molType=snp["molType"],
class_=snp["class"],
valid=snp["valid"],
avHet=snp["avHet"],
avHetSE=snp["avHetSE"],
func=snp["func"],
locType=snp["locType"],
weight=snp["weight"],
)
)
assay.cached_sequences.append(cached_seq)
Session.commit()
return sequences
def cut(self):
left_padding = self.form_result['left_padding']
right_padding = self.form_result['right_padding']
enzyme = Session.query(Enzyme).get(self.form_result['enzyme'])
cutseq = enzyme.cutseq
# TODO change to single amplicon?
if self.form_result['assay_id']:
assay = Session.query(SequenceGroup).get(self.form_result['assay_id'])
amplicon_tuples = pcr_sequences_snps_for_group(assay, padding_pos5=left_padding, padding_pos3=right_padding)
sequences = []
for amp, pseqs in amplicon_tuples:
sequences.extend(pseqs)
else:
manual_seq = PCRSequence(SimpleGenomeSequence(0, 0, len(self.form_result['positive_sequence'])-1, '+',
full_sequence=self.form_result['positive_sequence']))
manual_seq.snps = []
sequences = [manual_seq]
# TODO: this is arbitrary
location_cut_data = self.__enzyme_cut_locations(sequences[0], [enzyme])
# TODO support multiple sequences, somehow.
pos_seq = sequences[0].merged_positive_sequence
total_width = len(pos_seq)
re_width_pct = 100*float(len(cutseq))/total_width
enzyme_cut_data = location_cut_data[self.form_result['enzyme']]
positive_matches = []
negative_matches = []
return_dict = {}
snp_dict = dict([(s['name'], s) for s in sequences[0].snps])
# keys here are going to be amplicon_cuts, left_cuts and right_cuts, left_cut
for k, v in sorted(enzyme_cut_data.items()):
blank, original_positives, original_negatives = v[0]
snp_positives = []
snp_negatives = []
cancel_positives = []
cancel_negatives = []
for cuts in v[1:]:
snp_name, shifted_positives, shifted_negatives = cuts
snp = snp_dict[snp_name]
# TODO: this is an oversimplification but will probably only result
# in a shift in a particular restriction site
#
# TODO: I think this is sketchy right at the edges, needs to be tested. (> vs >=, etc)
# TODO: the code could stand to be more compact as well.
if len(shifted_positives) > len(original_positives):
found = False
for start, end, strand in shifted_positives:
if (snp['chromEnd'] >= pos_seq.start+start-1 and snp['chromEnd'] <= pos_seq.start+end) or \
(snp['chromStart'] >= pos_seq.start+start-1 and snp['chromStart'] <= pos_seq.start+end):
snp_positives.append((start, end, strand))
found = True
if not found:
pass
#raise Exception, "ERROR: additional positive strand restriction site not found by analyzing SNPs"
elif len(shifted_positives) < len(original_positives):
found = False
for start, end, strand in original_positives:
if (snp['chromEnd'] >= pos_seq.start+start-1 and snp['chromEnd'] <= pos_seq.start+end) or \
(snp['chromStart'] >= pos_seq.start+start-1 and snp['chromStart'] <= pos_seq.start+end) or \
(pos_seq.start+start-1 >= snp['chromStart'] and pos_seq.start+end <= snp['chromEnd']):
if (start, end, strand) not in cancel_positives:
cancel_positives.append((start, end, strand))
found = True
if not found:
pass
#raise Exception, "ERROR: cancelled positive strand restriction site not found by analyzing SNPs"
if len(shifted_negatives) > len(original_negatives):
# find where the new snp is
found = False
for start, end, strand in shifted_negatives:
if (snp['chromEnd'] >= pos_seq.end-(end+1) and snp['chromEnd'] <= pos_seq.end-start) or \
(snp['chromStart'] >= pos_seq.end-(end+1) and snp['chromStart'] <= pos_seq.end-start):
snp_negatives.append((start, end, strand))
found = True
if not found:
pass
# insertion screws you here.
#raise Exception, (snp['chromEnd'], snp['chromStart'], pos_seq.end, pos_seq.end-(shifted_negatives[0][1]+1), pos_seq.end-(shifted_negatives[0][0]))
#raise Exception, "ERROR: additional negative strand restriction site not found by analyzing SNPs"
elif len(shifted_negatives) < len(original_negatives):
found = False
for start, end, strand in original_negatives:
if (snp['chromEnd'] >= pos_seq.end-(end+1) and snp['chromEnd'] <= pos_seq.end-start) or \
(snp['chromStart'] >= pos_seq.end-(end+1) and snp['chromStart'] <= pos_seq.end-start) or \
(pos_seq.end-(end+1) >= snp['chromStart'] and pos_seq.end-start <= snp['chromEnd']):
if (start, end, strand) not in cancel_negatives:
cancel_negatives.append((start, end, strand))
found = True
if not found:
pass
#raise Exception, "ERROR: cancelling negative strand restriction site not found by analyzing SNPs"
for tup in cancel_positives:
original_positives.remove(tup)
for tup in cancel_negatives:
original_negatives.remove(tup)
return_dict[k] = len(original_positives) + len(cancel_positives) \
+ len(original_negatives) + len(cancel_negatives) \
+ len(snp_positives) + len(snp_negatives)
for start, end, strand in original_positives:
positive_matches.append({'offset': start, 'pos': '%s%%' % (start*100.0/total_width), 'class': 'stable_re_site'})
for start, end, strand in original_negatives:
negative_matches.append({'offset': start, 'pos': '%s%%' % (100-(start*100.0/total_width)-re_width_pct), 'class': 'stable_re_site'})
for start, end, strand in snp_positives:
positive_matches.append({'offset': start, 'pos': '%s%%' % (start*100.0/total_width), 'class': 'snp_re_site'})
for start, end, strand in snp_negatives:
negative_matches.append({'offset': start, 'pos': '%s%%' % (100-(start*100.0/total_width)-re_width_pct), 'class': 'snp_re_site'})
for start, end, strand in cancel_positives:
positive_matches.append({'offset': start, 'pos': '%s%%' % (start*100.0/total_width), 'class': 'snp_cancel_re_site'})
for start, end, strand in cancel_negatives:
negative_matches.append({'offset': start, 'pos': '%s%%' % (100-(start*100.0/total_width)-re_width_pct), 'class': 'snp_cancel_re_site'})
return_dict['positive_cuts'] = positive_matches
return_dict['negative_cuts'] = negative_matches
return_dict['re_width_pct'] = "%s%%" % re_width_pct
# future out amplicon position
amplicon_start = left_padding
amplicon_end = len(pos_seq) - (right_padding+1)
left_offsets = [match['offset'] for match in positive_matches if match['offset'] < (amplicon_start - len(cutseq))]
if left_offsets:
rightmost_left = max(left_offsets)+len(cutseq)
else:
rightmost_left = None
right_offsets = [match['offset'] for match in positive_matches if match['offset'] > amplicon_end]
if right_offsets:
leftmost_right = min(right_offsets)
else:
leftmost_right = None
amplicon_cuts = [match for match in positive_matches if match['offset'] >= amplicon_start and match['offset'] <= amplicon_end]
# todo: bug if the cutters are asymmetric and the negative cutsite is shorter (redmine 669)
if rightmost_left is not None and leftmost_right is not None and len(amplicon_cuts) == 0:
inner_len = leftmost_right - rightmost_left
inner_seq = pos_seq.sequence[rightmost_left:leftmost_right]
inner_gc = gc_content(inner_seq)
left_offset = amplicon_start - rightmost_left
right_offset = leftmost_right - amplicon_end
return_dict['fragment'] = {'len': inner_len,
'loff': left_offset,
'roff': right_offset,
'gc': "%.2f%%" % (inner_gc*100)}
return return_dict
