def test_haplotype_3070():
sites = str2sites('16093C 16183d 16184d 16191.1T 16191.2T 16270T')
seq = sites2seq(sites, region=range(16000,16570))
rts = seq2sites(seq) # rts: round trip sites
print 'EXP: %s' % sites
print 'OBS: %s' % rts
assert sites == rts
def test_haplotype_4827():
sites = str2sites('16172C 16183d 16193.1C 16193.2C 16223T 16320T')
seq = sites2seq(sites, region=range(16000,16570))
rts = seq2sites(seq) # rts: round trip sites
print 'EXP: %s' % sites
print 'OBS: %s' % rts
assert sites == rts
def test_haplotype_2236():
sites = str2sites('16126C 16163G 16185.1T 16185.2T 16189d 16294T 16519C')
seq = sites2seq(sites, region=range(16000,16570))
rts = seq2sites(seq) # rts: round trip sites
print 'EXP: %s' % sites
print 'OBS: %s' % rts
assert sites == rts
def test_haplotype_2911():
sites = str2sites('16051G 16129C 16182d 16183d 16193.1C 16193.2C 16362C 16519C')
seq = sites2seq(sites, region=range(16000,16570))
rts = seq2sites(seq) # rts: round trip sites
print 'EXP: %s' % sites
print 'OBS: %s' % rts
assert sites == rts
drop = ['San_43', 'San_67', 'tzbg040', 'tzdt045', 'tzhz108', 'tzhz130', 'tzhz131']
hids = []
seqs = []
for e in data:
name_parts = e['name'].split()
if name_parts[0] not in drop:
hids.append(name_parts[0])
seqs.append(e['sequence'])
## Validate
passed_validation = True
for i in range(len(seqs)):
mysites = seq2sites(seqs[i])
myseq = translate(sites2seq(mysites, region), None, '-')
if not seqs[i] == myseq:
passed_validation = False
print i, hids[i]
if passed_validation:
with open('processed.csv', 'w') as f:
for i in range(len(seqs)):
mysites = ' '.join([str(x) for x in seq2sites(seqs[i])])
origid = hids[i]
prefix = metadata.ix[origid[:4],'NewPrefix']
num = origid[4:].split('_')[0].zfill(3)
newid = prefix + num
f.write('%s,%s,%s\n' % (newid, origid, mysites))
from utils import *
## load metadata
metadata = pd.read_csv('metadata.csv', index_col=0)
region = range2region(metadata.ix[0, 'SeqRange'])
ff = fasta('krings1999.fasta', 'r')
data = ff.readentries()
ff.close()
hids = []
sites = []
for entry in data:
hids.append(entry['name'])
sites.append(seq2sites(entry['sequence']))
# some of the sequences are short. Many are just missing a base or two
# from the beginning or the end - will keep those
# a few are missing large chunks of the end, so will drop those
ok = [41, 42, 43, 44, 54, 55, 83, 85, 89, 136, 137, 157, 178]
skip = [8, 123, 124]
# validate
passed_validation = True
for i in range(len(sites)):
if i in ok or i in skip:
pass
else:
seq1 = data[i]['sequence'].upper()
if e['sequence'].count('n') > 10:
print 'skipping isolate %s' % name[3]
else:
hids.append(name[3])
seqs.append(e['sequence'])
if name[3] == '63':
pops.append('Yoruban')
else:
for f in e['features']:
if f[0] == 'source':
for f2 in f[1]:
if isinstance(f2, tuple):
if f2[0] == 'note':
pops.append(f2[1])
if 'complete' in e['definition'] or name[3] in ['3', '65']:
sites.append(seq2sites(e['sequence']))
else:
found = False
for f in e['features']:
if f[0] == 'misc_feature' and f[1][1][1] == 'segment 1':
c = int(f[1][0].split('..')[1]) - 1
s = e['sequence']
sites.append(seq2sites(s[:c]) + seq2sites(s[c:]))
found = True
if not found:
print 'problem with isolate %s' % name[3]
# Vigilant GenBank data have variable sequence lengths
# normalize all sites to specified range in the metadata file
mysites = []
hids1 = []
groups1 = []
seqs1 = []
sites1 = []
for e in entries1:
hid = getisolate(e)
if hid is not None:
hids1.append(hid)
n = getnote(e)
if n is not None:
parts = n.split()
groups1.append(parts[-1])
seqs1.append(e['sequence'])
sites1.append(seq2sites(e['sequence']))
entries2 = read_genbank('hvr2.gb', what='filename')
hids2 = []
groups2 = []
seqs2 = []
sites2 = []
for e in entries2:
hid = getisolate(e)
if hid is not None:
hids2.append(hid)
n = getnote(e)
if n is not None:
parts = n.split()
for e in data1:
k = e['name'].split()[0]
hvr1[k] = e['sequence']
for e in data2:
k = e['name'].split()[0]
hvr2[k] = e['sequence']
# retaining only those with both hvr1 and hvr2
k1 = set(hvr1.keys())
k2 = set(hvr2.keys())
hids = list(k1 & k2)
sites = {}
for k in hids:
sites[k] = seq2sites(hvr1[k])
sites[k] = sites[k] + seq2sites(hvr2[k])
## Validate
passed_validation = True
for i in range(len(sites)):
hid = hids[i]
seq1 = translate(sites2seq(sites[hid], region1), None, '-')
seq2 = translate(sites2seq(sites[hid], region2), None, '-')
if not seq1 == hvr1[hid] and seq2 == hvr2[hid]:
passed_validation = False
print i, hids[i]
if passed_validation:
counter = {}
def process_seq2sites(form):
"""Process data submitted in seq2sites form"""
# submission validation and error reporting
problems = []
valid = True
# first, just assume whatever is in the textarea is the submission
# even if that may be nothing
content = form.cleaned_data['query']
# then check to see if a file was supplied, and if so, replace the
# previously assumed content with the file data
if form.cleaned_data['file'] is not None:
if form.cleaned_data['file'].multiple_chunks():
pass
# error - return with error
content = form.cleaned_data['file'].read()
# clear off any trailing whitespace
content.strip()
# make sure something was submitted
if len(content) == 0:
valid = False
return HttpResponseRedirect(reverse('seq2sites'))
# determine format
format = None
if content.startswith('>'):
format = 'fasta'
else:
format = 'single_seq'
# pull names and sequence out of submitted content
names = []
seqs = []
if format == 'fasta':
try:
fnames = []
fseqs = []
for entry in fasta(content, 's'):
fnames.append(entry['name'])
fseqs.append(RE_NON_IUPAC.sub('', entry['sequence'].upper()))
names = fnames
seqs = fseqs
except:
valid = False
problems.append('There was an error in the FASTA format')
else:
names = ['']
seqs = [RE_NON_IUPAC.sub('', content.upper())]
# enforce limits for multisequence submissions
if format == 'fasta':
if len(seqs) > MAX_SEQS:
valid = False
problems.append('Too many sequences submitted; current maximum allowed is %d' % MAX_SEQS)
if valid:
result_lines = []
sites_by_line = []
for seq in seqs:
try:
sites = seq2sites(seq)
sites_by_line.append(sites)
result_lines.append(sites2str(sites))
except Exception, e:
result_lines.append('There was an error: %s' % e)
results = list(Result(x,y) for x,y in zip(names,result_lines))
c = Context({'results': results})
s = "%s%s" % m.groups()
elif ".1" in s:
s = s + "C"
elif ".2" in s:
h2f.append(s[:-1] + "1C")
s = s + "C"
h2f.append(s)
hvr2[i] = " ".join(h2f)
## Validate
passed_validation = True
for i in range(len(freq)):
curr_sites = str2sites(hvr1[i], add16k=True)
seq = translate(sites2seq(curr_sites, region1), None, "-")
mysites = seq2sites(seq)
if not mysites == curr_sites:
myseq = translate(sites2seq(mysites, region1), None, "-")
if not seq == myseq:
passed_validation = False
print i, "hvr1"
curr_sites = str2sites(hvr2[i])
seq = translate(sites2seq(curr_sites, region2), None, "-")
mysites = seq2sites(seq)
if not mysites == curr_sites:
myseq = translate(sites2seq(mysites, region2), None, "-")
if not seq == myseq:
passed_validation = False
print i, "hvr2"
ff = fasta('knight2003.fasta', 'r')
data = ff.readentries()
ff.close()
hids = []
seqs = []
sites = []
for e in data:
hids.append(e['name'].split()[0])
seqs.append(e['sequence'])
m = re.search(r'GATCACA', e['sequence'])
cut = m.start()
seq1 = e['sequence'][:cut]
seq2 = e['sequence'][cut:]
sites.append(seq2sites(seq1) + seq2sites(seq2))
## Validate
passed_validation = True
for i in range(len(sites)):
seq = translate(sites2seq(sites[i], region1), None, '-') + translate(sites2seq(sites[i], region2), None, '-')
if not seq == seqs[i]:
passed_validation = False
print i, hids[i]
if passed_validation:
counter = {}
for k in metadata.index:
counter[k] = 0
with open('processed.csv', 'w') as f:
sites.append(" ".join(parts[3:]))
## Validate
passed_validation = True
# there are sites in the source table that are not actual variant sites
# sequence 9 (index 8) has 263A as a variant
# sequence 12 (index 11) has 16223C as a variant
not_polys = [Polymorphism(263, 0, "A"), Polymorphism(16223, 0, "C")]
for i in range(len(sites)):
curr_sites = sites[i]
curr_polys = [x for x in str2sites(curr_sites) if x not in not_polys]
cseq1 = sites2seq(curr_sites, region1)
cseq2 = sites2seq(curr_sites, region2)
mysites1 = seq2sites(cseq1)
mysites2 = seq2sites(cseq2)
mysites = mysites1 + mysites2
if not mysites == curr_polys:
passed_validation = False
print iids[i]
if passed_validation:
counters = [1] * 2
with open("processed.csv", "w") as f:
for i in range(len(sites)):
curr_sites = str2sites(sites[i])
mysites = [x for x in curr_sites if x not in not_polys]
mysites = " ".join([str(x) for x in mysites])
for j in range(2):
if counts[i, j] > 0:
def new_query(self, seq, label='Query'):
all_polymorphisms = {}
defining_polymorphisms = {}
return MotifQuery(defining_polymorphisms=seq2sites(seq), label=label)
ff = fasta('non_2011.fasta', 'r')
data = ff.readentries()
ff.close()
hids = []
seqs = []
sites = []
# four sequences are shorter than all the rest, will drop them
for e in data:
if len(e['sequence']) > 350:
hids.append(e['name'].split()[0])
seqs.append(e['sequence'])
sites.append(seq2sites(e['sequence']))
## Validate
passed_validation = True
for i in range(len(sites)):
hid = hids[i]
key = hid[:2]
region = range2region(metadata.ix[key, 'SeqRange'])
seq = translate(sites2seq(sites[i], region), None, '-')
if not seq == seqs[i]:
passed_validation = False
print i, hids[i]
if passed_validation:
counter = {}
sites = []
with open('defilippo2010.csv', 'rU') as f:
reader = csv.reader(f)
reader.next() # skip past header
for row in reader:
hids.append(row[0])
groups.append(row[1])
sites.append(str2sites(row[3]))
## Validate
passed_validation = True
for i in range(len(sites)):
seq = sites2seq(sites[i], region)
mysites = seq2sites(seq)
if not mysites == sites[i]:
myseq = translate(sites2seq(mysites, region), None, '-')
if not seq == myseq:
passed_validation = False
print i, hids[i]
counter = {}
for k in metadata.index:
counter[k] = 0
if passed_validation:
with open('processed.csv', 'w') as f:
for i in range(len(groups)):
key = groups[i]
counter[key] = counter[key] + 1
for filename in ["barbieri2013a.fasta", "barbieri2013b.fasta", "barbieri2013c.fasta", "barbieri2014.fasta"]:
ff = fasta(filename, "r")
data = ff.readentries()
ff.close()
for e in data:
if e["sequence"].count("N") < 20:
hid = e["name"].split()[4].upper()
if hid in groups.index:
hids.append(hid)
seqs.append(e["sequence"])
sites = []
for s in seqs:
sites.append(seq2sites(s, ambig_cutoff=20))
## Validate
passed_validation = True
for i in range(len(sites)):
hid = hids[i]
seq = translate(sites2seq(sites[i], region), None, "-")
if not seq == seqs[i]:
# some sequences have N in position 308 and this doesn't parse properly in my converter
# not going to worry about it because will always drop the variants in this region
if not seq[:305] == seqs[i][:305] and seq[-16260:] == seqs[i][-16260:]:
passed_validation = False
print i, hids[i]
if passed_validation:
