def encl_sam_genotype(sam_path, arg_dict):
exp_dir = arg_dict['exp_dir']
read_len = arg_dict['read_len']
motif = arg_dict['motif']
score_dict = { 'match': 3, \
'mismatch': -1, \
'gap': -3}
verbose = False
pre, post = extract_pre_post_flank(exp_dir, read_len)
nCopy_dict = {}
total_count = 0
with open(sam_path, 'r') as encl_handle:
for record in csv.reader(encl_handle, dialect='excel-tab'):
if record[0][0] != '@':
sample = record[9]
nCopy, pos, score = expansion_aware_realign(
sample, pre, post, motif, score_dict, verbose)
if nCopy not in nCopy_dict:
nCopy_dict[nCopy] = 1
else:
nCopy_dict[nCopy] = nCopy_dict[nCopy] + 1
total_count = total_count + 1
nCopy_list = nCopy_dict.keys()
freq_list = []
for nCopy in nCopy_list:
freq_list.append(float(nCopy_dict[nCopy]) / float(total_count))
return nCopy_list, freq_list
print TLEN
nc_col = 'nc:i:' + str(0)
ps_col = 'ps:i:' + str(0)
sc_col = 'sc:i:' + str(0)
rc_col = 'rc:Z:' + 'unknown'
is_col = 'is:i:' + str(int(np.abs(TLEN)))
out_sam_handle.write(
'\t'.join(row + [nc_col, ps_col, sc_col, rc_col, is_col]) +
'\n')
elif (RNAME == chrom and POS <= locus_start - read_len and RNEXT == '=' and PNEXT <= locus_start) or \
(RNAME == chrom and POS >= locus_end and RNEXT == '=' and PNEXT >= locus_end):
# ignoring read pairs that are both before or after STR region.
pass
else:
# We realign first read to check if it's pre or post flanking
nCopy, pos, score = expansion_aware_realign(
SEQ, pre, post, motif, score_dict, verbose)
nCopy_rev, pos_rev, score_rev = expansion_aware_realign(
reverse_strand(SEQ), pre, post, motif, score_dict, verbose)
if score_rev > score:
nCopy = nCopy_rev
pos = pos_rev
score = score_rev
SEQ = reverse_strand(SEQ)
read_class = classify_realigned_read(SEQ, motif, pos, nCopy,
score, score_dict,
read_len, margin, verbose)
# We're allowing flanking from both sides, but one side should be mapped to the correct location for this to work.
# i.e. we don't rescue the case that both reads are flanking, and both are mapped to a different locus.
nc_col = 'nc:i:' + str(nCopy)
ps_col = 'ps:i:' + str(pos)
sc_col = 'sc:i:' + str(score)
verbose = False
margin = 2
chrom, locus_start, locus_end = extract_locus_info(locus_bed)
with open(in_sam, 'r') as in_sam_handle:
for record in in_sam_handle:
if record[0] == '@':
# out_sam_handle.write(record)
pass
else:
row = record.split()
if (row[2] == chrom and int(row[3]) >= locus_start - read_len and
int(row[3]) <= locus_end): # Reads mapped within region
sample = row[9]
nCopy, pos, score = expansion_aware_realign(
sample, pre, post, motif, score_dict, verbose)
read_class = classify_realigned_read(sample, motif, pos, nCopy,
score, score_dict,
read_len, margin, verbose)
print '>> ', read_class, ':'
print '>> nCopy =', nCopy
print '>> score =', score
print '>> pos =', pos
print sample
print
# if (row[2] == chrom and row[6] == '=') or (row[6] == chrom): # checking correct chrom for mate
# if int(row[7]) <= locus_start: # checking if mate is before STR region
# if row[2] != chrom:
# out_sam_handle.write(record)
# print row[9]