def get_cigar(s1, s2):
import parasail
if kind == "local":
result = parasail.sw_trace(s1, s2, 101, 10, parasail.pam100)
elif kind == "semi-global":
result = parasail.sg_trace(s1, s2, 101, 10, parasail.pam100)
elif kind == "global":
result = parasail.nw_trace(s1, s2, 101, 10, parasail.pam100)
else:
raise ValueError(
"The kind of alignment must be global, semi-global, or local.")
output = []
for i in result.cigar.seq:
#print(result.cigar.decode_len(i), result.cigar.decode_op(i).decode())
output += [
result.cigar.decode_len(i),
result.cigar.decode_op(i).decode()
]
for i in output[:0]:
float(i)
total_base = sum(output[:0])
# I'll convert to the canonical way of showing these things (e.g. SAMFILES - pysam)
# CIGAR type
ct = output[1::2]
# CIGAR length
cl = output[::2]
output = list(zip(ct, cl))
return output
def test1():
p = parasail.ssw_init("asdf", parasail.blosum62, 1)
r = parasail.ssw_profile(p, "asdf", 10, 1)
print(p.s1)
print(p.s1Len)
print(r.cigarLen)
print(r.cigar[0])
r = parasail.sw_trace("asdf", "asdf", 10, 1, parasail.blosum62)
c = r.cigar
print(c.len)
print(c.seq[0])
print(c.decode)
p = parasail.profile_create_8("asdf", parasail.blosum62)
r = parasail.sw_trace_striped_profile_8(p, "asdf", 10, 1)
c = r.cigar
print(c.len)
print(c.seq[0])
r = parasail.sw_trace("asdf", "asdf", 10, 1, parasail.blosum62)
print(r.query)
print(r.ref)
def test1(self):
p = parasail.ssw_init("asdf", parasail.blosum62, 1)
r = parasail.ssw_profile(p, "asdf", 10, 1)
print(p.s1)
print(p.s1Len)
print(r.cigarLen)
print(r.cigar[0])
r = parasail.sw_trace("asdf", "asdf", 10, 1, parasail.blosum62)
c = r.cigar
print(c.len)
print(c.seq[0])
print(c.decode)
p = parasail.profile_create_8("asdf", parasail.blosum62)
r = parasail.sw_trace_striped_profile_8(p, "asdf", 10, 1)
c = r.cigar
print(c.len)
print(c.seq[0])
r = parasail.sw_trace("asdf", "asdf", 10, 1, parasail.blosum62)
print(r.query)
print(r.ref)
def _get_hit_parasail(vj,
all_genes,
organism,
ab,
blast_seq,
user_matrix=ps.matrix_create("ACGT", 3, -5)):
"""
The workhorse function that can be inserted into the VJ loop in parse_unpaired_dna_sequence_blastn()
It returns objects that mimic the returns of functions:
parse_blast_alignments()
get_all_hits_with_evalues_and_scores()
:param vj:
:param all_genes:
:param organism:
:param ab:
:param blast_seq:
:param user_matrix:
:return: 2-part dictionary containing "hits" and "hits_scores
{"hits" : {"tmp": ParasailMatch.instance}, "hits_scores" : [(id,score,evalue),((id,score,evalue))]}
"hits" returns a ParasailMatch instance that mimics the attributes of a BlastMatch instance
"hits_scores" returns a list of 3-part tuples withm, 0: hit_id, 1: alignment score, 2: evalue approximation
"""
ids = _get_ids_by_org_chain_region(organism=organism,
chain=ab,
region=vj,
d=all_genes)
# seqs is a list of tuples
# 0: hit_id,
# 1: hit_seq (full length reference seq)
# 2: strand (-1 = rev comp)
seqs = _get_sequence_tuples_from_ids(ids=ids,
organism=organism,
d=all_genes)
scores = []
user_matrix = ps.matrix_create("ACGT", 1, -3) # -5
for i in range(len(seqs)):
# smith-waterman alignment implemented using parasail
s = ps.sw_trace(
s1=blast_seq,
s2=seqs[i][1],
extend=2, #3
open=5, #5
matrix=user_matrix)
scores.append({
'score': s.score,
'parasail_result': s,
'query_seq': blast_seq,
'hit_id': seqs[i][0],
'hit_seq': seqs[i][1],
'h_strand': seqs[i][2],
'q_strand': 1
})
# sort parasail results from highest to lowest alignment score
scores = sorted(scores, key=lambda x: x['score'], reverse=True)
id_score_evalue = [(s['hit_id'], s['score'],
_evalue_aproximation(s['score'])) for s in scores]
# add alignment start positions for the highest scoring alignment
scores[0]["parasail_result"].get_traceback()
scores[0]["q_seq"] = scores[0]["parasail_result"]._traceback.query
scores[0]["h_seq"] = scores[0]["parasail_result"]._traceback.ref
scores[0]["comp"] = scores[0]["parasail_result"]._traceback.comp
scores[0]["q_start"] = _q_start(q_seq=scores[0]["q_seq"],
query_seq=blast_seq)
scores[0]["q_stop"] = _q_stop(q_seq=scores[0]["q_seq"],
query_seq=blast_seq)
scores[0]["h_start"] = _h_start(h_seq=scores[0]["h_seq"],
hit_seq=scores[0]["hit_seq"],
h_strand=scores[0]["h_strand"])
scores[0]["h_stop"] = _h_stop(h_seq=scores[0]["h_seq"],
hit_seq=scores[0]["hit_seq"],
h_strand=scores[0]["h_strand"])
scores[0]["identities"] = _identities(scores[0]["comp"])
# add q2hmap for the highest scoring alignment
q2hmap = _create_q2hmap(q_seq=scores[0]["q_seq"],
h_seq=scores[0]["h_seq"],
q_start=scores[0]['q_start'],
h_start=scores[0]['h_start'],
q_strand=scores[0]['q_strand'],
h_strand=scores[0]['h_strand'])
phony_evalue_must_update_function = _evalue_aproximation(
scores[0]['score'])
# bm2 is going to replace teh BlastMatch instance passed by parse_blast_alignments()
# we only produce it for the top scoring hit scores[0], but the code is written,
# so that we could produce ParasailMatches in a loop
bm2 = ParasailMatch(query_id="tmp", hit_id=scores[0]['hit_id'])
bm2.evalue = phony_evalue_must_update_function #! SHOULD UPDATE
bm2.identities = scores[0]["identities"] # percent identities out of 100
bm2.h_start = scores[0]['h_start'] # 0-indexed
bm2.h_stop = scores[0]['h_stop']
bm2.h_strand = scores[0]['h_strand']
bm2.h_align = scores[0]['h_seq']
bm2.q_start = scores[0]['h_start']
bm2.q_stop = scores[0]['h_stop']
bm2.q_strand = scores[0]['q_strand']
bm2.q_align = scores[0]['q_seq']
bm2.middleseq = scores[0]['comp']
bm2.q2hmap = q2hmap # q2hmap ## 0-indexed numbering wrt to fullseq
bm2.valid = "True" # valid IF WHAT?
bm2.frame = 'NA'
# results are meant to mimic the outputs in prior functions from blast version:
# hits = parse_blast_alignments( blast_tmpfile+'.blast', evalue_threshold, identity_threshold )
# hits_scores = get_all_hits_with_evalues_and_scores( blast_tmpfile+'.blast' ) ## id,bitscore,evalue
results = {"hits": {"tmp": bm2}, "hits_scores": id_score_evalue}
return (results)
def test4():
parasail.set_case_sensitive(True)
matrix = parasail.matrix_create("ACGT", 2, 1)
result = parasail.sw_trace("ACGT", "AcgT", 10, 1, matrix)
traceback = result.traceback
print_traceback_attributes(traceback)
def test22():
matrix = parasail.matrix_create("ACGTacgt", 2, 1, True)
result = parasail.sw_trace("ACGT", "AcgT", 10, 1, matrix)
traceback = result.get_traceback(case_sensitive=True)
print_traceback_attributes(traceback)
def test21():
matrix = parasail.matrix_create("ACGTacgt", 2, 1, True)
result = parasail.sw_trace("ACGT", "AcgT", 10, 1, matrix)
traceback = result.traceback
print_traceback_attributes(traceback)
def test2():
result = parasail.sw_trace("asdf", "asdf", 10, 1, parasail.blosum62)
cigar = result.cigar
print_cigar_attributes(cigar)
def test3():
result = parasail.sw_trace("asdf", "asdf", 10, 1, parasail.blosum62)
traceback = result.traceback
print_traceback_attributes(traceback)
def test2():
result = parasail.sw_trace("asdf","asdf",10,1,parasail.blosum62)
cigar = result.cigar
print_cigar_attributes(cigar)
def test3():
result = parasail.sw_trace("asdf","asdf",10,1,parasail.blosum62)
traceback = result.traceback
print_traceback_attributes(traceback)
def test4():
parasail.set_case_sensitive(True)
matrix = parasail.matrix_create("ACGT", 2, 1)
result = parasail.sw_trace("ACGT","AcgT",10,1,matrix)
traceback = result.traceback
print_traceback_attributes(traceback)
def test22():
matrix = parasail.matrix_create("ACGTacgt", 2, 1, True)
result = parasail.sw_trace("ACGT","AcgT",10,1,matrix)
traceback = result.get_traceback(case_sensitive=True)
print_traceback_attributes(traceback)
def test21():
matrix = parasail.matrix_create("ACGTacgt", 2, 1, True)
result = parasail.sw_trace("ACGT","AcgT",10,1,matrix)
traceback = result.traceback
print_traceback_attributes(traceback)
def test3(self):
result = parasail.sw_trace("asdf", "asdf", 10, 1, parasail.blosum62)
print_traceback_attributes(result.traceback)
def test2(self):
result = parasail.sw_trace("asdf", "asdf", 10, 1, parasail.blosum62)
print_cigar_attributes(result.cigar)