def make_consensus(bam_file, ref_file, freebayes_vcf):
''':retrurn str'''
refs = list(SeqIO.parse(ref_file, 'fasta'))
df = bioframes.load_vcf(freebayes_vcf)
ids, raw_segment_dfs = zip(*df.groupby('CHROM'))
segment_dfs = map(fix_fb_df, raw_segment_dfs)
refs = sorted(refs, key=lambda x: x.id)
assert list(pluck_attr('id', refs)) == sorted(ids)
ref_seqs = (str(s.seq) for s in refs) #pluck_attr('seq', refs)
segment_dfs = sorted(segment_dfs, key=lambda x: x.CHROM.iloc[0])
def process(original_ref, df):
new_ref, off = reduce(swap_base, zip(df.POS, df.REF, df.ALT), (original_ref, 0))
#currently only handle inserts, > 0
offs, off_pos = df[df.OFF > 0].OFF, df[df.OFF > 0].POS
pileup_positions = zero_coverage_positions(bam_file, ref_file)
return gap_fill_ref(original_ref, new_ref, pileup_positions, offs, off_pos, off)
#return map(process, ref_seqs, segment_dfs)
return zip(sorted(ids), map(process, ref_seqs, segment_dfs))
def _write_all_lines(self, cr, uid, ids, vals_to_write, context=None):
stmt_line_obj = self.pool['bank.acc.rec.statement.line']
for stmt in self.browse(cr, uid, ids, context=context):
stmt_line_ids = map(
lambda x: x.id,
stmt.credit_move_line_ids + stmt.debit_move_line_ids)
stmt_line_obj.write(cr,
uid,
stmt_line_ids,
vals_to_write,
context=context)
return True
def onchange_account_id(self,
cr,
uid,
ids,
account_id,
ending_date,
suppress_ending_date_filter,
keep_previous_uncleared_entries,
context=None):
aml_obj = self.pool['account.move.line']
stmt_line_obj = self.pool['bank.acc.rec.statement.line']
val = {
'value': {
'credit_move_line_ids': [],
'debit_move_line_ids': []
}
}
if not account_id:
return val
for statement in self.browse(cr, uid, ids, context=context):
cr.execute(
'''
UPDATE account_move_line
SET draft_assigned_to_statement=False,
cleared_bank_account=False,
bank_acc_rec_statement_id=NULL
WHERE bank_acc_rec_statement_id=%s''', (statement.id, ))
cr.execute(
'''
DELETE FROM bank_acc_rec_statement_line
WHERE statement_id=%s''', (statement.id, ))
# Apply filter on move lines to allow
# 1. credit and debit side journal items in posted
# state of the selected GL account
# 2. Journal items which are not assigned to
# previous bank statements
# 3. Date less than or equal to ending date provided the
# 'Suppress Ending Date Filter' is not checked
# get previous uncleared entries
domain = [('account_id', '=', account_id),
('move_id.state', '=', 'posted'),
('cleared_bank_account', '=', False)]
if not keep_previous_uncleared_entries:
domain += [('draft_assigned_to_statement', '=', False)]
if not suppress_ending_date_filter:
domain += [('date', '<=', ending_date)]
if keep_previous_uncleared_entries:
keep = curry(self.is_stmt_done, cr, uid, context=context)
else:
keep = lambda x: True
aml_search = curry(aml_obj.search, cr, uid, context=context)
aml_browse = curry(aml_obj.browse, cr, uid, context=context)
is_credit = (lambda l: '%s_move_line_ids' %
(l['type'] == 'cr' and 'credit' or 'debit'))
val['value'].update(
pipe(domain, aml_search, aml_browse, filter(keep),
map(self._get_move_line_write), groupby(is_credit)))
val['value'].update({
'starting_balance':
self._get_starting_balance(cr,
uid,
ids and ids[0] or 0,
account_id,
context=context)
})
return val
def zero_coverage_positions(bam_file, ref_file):
pileup = sh.samtools('mpileup', bam_file, f=ref_file, _iter=True)
return map(compose(int, second, unicode.split), pileup)
def make_dict(classes):
return dict(zip(map(call('__name__'), classes, classes)))
# and the VCFs have multiple references. need to group by reference
def fix_fb_df(df):
#Freebayes only ever reports one ALT?
df.ALT = df.ALT.apply(lambda x: x[0])
# the vcf library reports alts as _Substitution/whatever objects. extract the string.
df.REF, df.ALT = df.REF.apply(str), df.ALT.apply(str)
'''#TODO: this re-definition of ambiguous bases translates mult-base
sections (e.g. AC) into single base alts or something'''
ambiguous = ((df.AO / df.DP.apply(float)) < 0.8)
#have to use .loc for assignment or else get shallow copy warning
#NOTE: temporarily removed the amiguous base business.
#df.loc[ambiguous, 'ALT'] = list(map(get_degen, zip(df.loc[ambiguous].REF, df.loc[ambiguous].ALT)))
df['OFF'] = df.ALT.apply(len) - df.REF.apply(len)
return df
pluck_attr = lambda a, A: map(attrgetter(a), A)
#string_to_fasta = '>FreebayesConseunsus\n'.__add__
string_to_fasta = '>Freebayes {0}\n{1}'.format
def make_consensus(bam_file, ref_file, freebayes_vcf):
''':retrurn str'''
refs = list(SeqIO.parse(ref_file, 'fasta'))
df = bioframes.load_vcf(freebayes_vcf)
ids, raw_segment_dfs = zip(*df.groupby('CHROM'))
segment_dfs = map(fix_fb_df, raw_segment_dfs)
refs = sorted(refs, key=lambda x: x.id)
assert list(pluck_attr('id', refs)) == sorted(ids)
ref_seqs = (str(s.seq) for s in refs) #pluck_attr('seq', refs)
segment_dfs = sorted(segment_dfs, key=lambda x: x.CHROM.iloc[0])
def process(original_ref, df):
new_ref, off = reduce(swap_base, zip(df.POS, df.REF, df.ALT), (original_ref, 0))
