for x in e['features']:
if x[0] == 'source':
for y in x[1]:
if isinstance(y, tuple):
if y[0] == 'isolate':
return y[1]
return None
## load metadata
metadata = pd.read_csv('metadata.csv', index_col=0)
regionparts = metadata.ix[0, 'SeqRange'].split(';')
region1 = range2region(regionparts[0])
region2 = range2region(regionparts[1])
entries1 = read_genbank('hvr1.gb', what='filename')
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'])
from genbank import read_genbank
## load metadata
metadata = pd.read_csv('metadata.csv', index_col=0)
regionstr = metadata.ix[0,'SeqRange']
regionparts = regionstr.split(';')
region1 = range2region(regionparts[0])
region2 = range2region(regionparts[1])
counts = {}
with open('vigilant1991_counts.txt', 'rU') as f:
for line in f:
parts = line.strip().split(',')
counts[parts[0]] = int(parts[1])
entries = read_genbank('vigilant1991.txt', what='filename')
hids = []
pops = []
seqs = []
sites = []
for i in range(len(entries)):
e = entries[i]
name = e['definition'].split()
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':
# go through the genbank file, extract ids and population attributions
# drop the mandenka and yoruba (already in db from Lippold et al. 2014)
# and write out in fasta
import sys
from oldowan.fasta import fasta
sys.path.append('../../scripts')
from utils import *
from genbank import read_genbank
entries = read_genbank('barbieri2012.gb', what='filename')
def getpop(x):
source = x['features'][0][1]
for x in source:
if isinstance(x, tuple):
if x[0] == 'pop_variant':
return x[1]
def getid(x):
words = x['definition'].split()
return words[3]
ff = fasta('barbieri2012.fasta', 'w')
for i in range(len(entries)):
hid = getid(entries[i])
if not hid.startswith('MAN') and not hid.startswith('YOR'):
pop = getpop(entries[i])
seq = entries[i]['sequence'].upper()
newentry = {'name': "%s %s" % (hid, pop), 'sequence':seq}
