def parasail_alignment(s1,
s2,
match_score=2,
mismatch_penalty=-2,
opening_penalty=3,
gap_ext=1):
user_matrix = parasail.matrix_create("ACGT", match_score, mismatch_penalty)
result = parasail.sg_trace_scan_16(s1, s2, opening_penalty, gap_ext,
user_matrix)
if result.saturated:
print("SATURATED!", len(s1), len(s2))
result = parasail.sg_trace_scan_32(s1, s2, opening_penalty, gap_ext,
user_matrix)
print("computed 32 bit instead")
# difference in how to obtain string from parasail between python v2 and v3...
if sys.version_info[0] < 3:
cigar_string = str(result.cigar.decode).decode('utf-8')
else:
cigar_string = str(result.cigar.decode, 'utf-8')
s1_alignment, s2_alignment, cigar_tuples = cigar_to_seq(
cigar_string, s1, s2)
# print(result.score, len(s1), len(s2))
# print(s1_alignment)
# print(s2_alignment)
# print(cigar_string)
# sys.exit()
# print(dir(result))
# print(result.end_query, result.end_ref, result.len_query, result.len_ref, result.length, result.matches)
# print()
return s1_alignment, s2_alignment, cigar_string, cigar_tuples, result.score
def parasail_block_alignment(s1,
s2,
k,
match_id,
x_acc="",
y_acc="",
match_score=2,
mismatch_penalty=-2,
opening_penalty=5,
gap_ext=1,
ends_discrepancy_threshold=0):
user_matrix = parasail.matrix_create("ACGT", match_score, mismatch_penalty)
result = parasail.sg_trace_scan_16(s1, s2, opening_penalty, gap_ext,
user_matrix)
if result.saturated:
print("SATURATED!")
result = parasail.sg_trace_scan_32(s1, s2, opening_penalty, gap_ext,
user_matrix)
if sys.version_info[0] < 3:
cigar_string = str(result.cigar.decode).decode('utf-8')
else:
cigar_string = str(result.cigar.decode, 'utf-8')
s1_alignment, s2_alignment = cigar_to_seq(cigar_string, s1, s2)
# Rolling window of matching blocks
# k=15
# match_id = int(k*0.8) 1.0 - math.ceil(window_fraction)
match_vector = [
1 if n1 == n2 else 0 for n1, n2 in zip(s1_alignment, s2_alignment)
]
# print("".join([str(m) for m in match_vector]))
match_window = deque(match_vector[:k]) # initialization
current_match_count = sum(match_window)
aligned_region = []
if current_match_count >= match_id:
aligned_region.append(1)
else:
aligned_region.append(0)
for new_m_state in match_vector[k:]:
prev_m_state = match_window.popleft()
current_match_count = current_match_count - prev_m_state + new_m_state
match_window.append(new_m_state)
if current_match_count >= match_id:
aligned_region.append(1)
else:
aligned_region.append(0)
# print("".join([str(m) for m in aligned_region]))
# print("Aligned ratio (tot aligned/len(seq1):", sum(aligned_region)/float(len(s1)))
alignment_ratio = sum(aligned_region) / float(len(s1))
return (s1, s2, (s1_alignment, s2_alignment, alignment_ratio))
def parasail_block_alignment(s1,
s2,
k,
match_id,
match_score=2,
mismatch_penalty=-2,
opening_penalty=5,
gap_ext=1):
user_matrix = parasail.matrix_create("ACGT", match_score, mismatch_penalty)
result = parasail.sg_trace_scan_16(s1, s2, opening_penalty, gap_ext,
user_matrix)
if result.saturated:
print("SATURATED!", len(s1), len(s2))
result = parasail.sg_trace_scan_32(s1, s2, opening_penalty, gap_ext,
user_matrix)
print("computed 32 bit instead")
# difference in how to obtain string from parasail between python v2 and v3...
if sys.version_info[0] < 3:
cigar_string = str(result.cigar.decode).decode('utf-8')
else:
cigar_string = str(result.cigar.decode, 'utf-8')
s1_alignment, s2_alignment = help_functions.cigar_to_seq(
cigar_string, s1, s2)
# Rolling window of matching blocks
match_vector = [
1 if n1 == n2 else 0 for n1, n2 in zip(s1_alignment, s2_alignment)
]
match_window = deque(match_vector[:k]) # initialization
current_match_count = sum(match_window)
aligned_region = []
if current_match_count >= match_id:
aligned_region.append(1)
else:
aligned_region.append(0)
for new_m_state in match_vector[k:]:
prev_m_state = match_window.popleft()
current_match_count = current_match_count - prev_m_state + new_m_state
match_window.append(new_m_state)
if current_match_count >= match_id:
aligned_region.append(1)
else:
aligned_region.append(0)
# print("".join([str(m) for m in aligned_region]))
# print("Aligned ratio (tot aligned/len(seq1):", sum(aligned_region)/float(len(s1)))
alignment_ratio = sum(aligned_region) / float(len(s1))
return (s1, s2, (s1_alignment, s2_alignment, alignment_ratio))
def parasail_local(s1,
s2,
match_score=2,
mismatch_penalty=-2,
opening_penalty=3,
gap_ext=1):
user_matrix = parasail.matrix_create("ACGT", match_score, mismatch_penalty)
result = parasail.sw_trace_scan_16(s1, s2, opening_penalty, gap_ext,
user_matrix)
if result.saturated:
print("SATURATED!", len(s1), len(s2))
result = parasail.sg_trace_scan_32(s1, s2, opening_penalty, gap_ext,
user_matrix)
print("computed 32 bit instead")
# difference in how to obtain string from parasail between python v2 and v3...
if sys.version_info[0] < 3:
cigar_string = str(result.cigar.decode).decode('utf-8')
else:
cigar_string = str(result.cigar.decode, 'utf-8')
s1_alignment, s2_alignment, cigar_tuples = cigar_to_seq(
cigar_string, s1[result.cigar.beg_query:result.end_query],
s2[result.cigar.beg_ref:result.end_ref])
# print(result.traceback.ref)
# print(result.traceback.comp)
# print(result.traceback.query)
# print(result.score, len(s1), len(s2))
print("read", s1_alignment)
print("Rref", s2_alignment)
print(result.cigar.beg_query, result.end_query)
print(result.cigar.beg_ref, result.end_ref)
print(cigar_string)
# print(result.cigar.seq)
# sys.exit()
# print(dir(result))
# for attr, value in result.__dict__.items():
# print(attr, value)
# print(result.end_query, result.end_ref, result.len_query, result.len_ref, result.length, result.matches)
# print()
return s1_alignment, s2_alignment, cigar_string, cigar_tuples, result.score
def parasail_alignment(read,
reference,
x_acc="",
y_acc="",
match_score=2,
mismatch_penalty=-2,
opening_penalty=2,
gap_ext=1,
ends_discrepancy_threshold=0):
user_matrix = parasail.matrix_create("ACGT", match_score, mismatch_penalty)
result = parasail.sg_trace_scan_16(read, reference, opening_penalty,
gap_ext, user_matrix)
if result.saturated:
print("SATURATED!")
result = parasail.sg_trace_scan_32(read, reference, opening_penalty,
gap_ext, user_matrix)
if sys.version_info[0] < 3:
cigar_string = str(result.cigar.decode).decode('utf-8')
else:
cigar_string = str(result.cigar.decode, 'utf-8')
read_alignment, ref_alignment = cigar_to_seq(cigar_string, read, reference)
return read_alignment, ref_alignment
def aln_nucleotides(seq1, name1, seq2, name2):
result = parasail.sg_trace_scan_32(seq1, seq2, 10, 1, parasail.nuc44)
return construct_psl(name1, name2, result)
def aln_proteins(seq1, name1, seq2, name2):
result = parasail.sg_trace_scan_32(seq1, seq2, 10, 1, parasail.blosum62)
return construct_psl(name1, name2, result)