def run_minimus(fasta,
outroot=None,
restore_singletons=True,
contig_prefix='',
qual=None):
'''given a fasta file and an optional output root (otherwise use fasta base)
generates an assembly using minimus from the amos package
assembly saved as <outroot>.minimus.fasta
optionally restores singleton reads in <outroot>.all.fasta
if restore_singletons=True, returns path to .all.fasta, otherwise returns path to minimus.fasta'''
if outroot is None:
outroot = fasta.rsplit('.', 1)[0]
if qual:
print >> sys.stderr, 'qualities invoked (%s)' % qual
os.system('toAmos -s %s -q %s -o %s.afg' % (fasta, qual, outroot))
else:
os.system('toAmos -s %s -o %s.afg' % (fasta, outroot))
os.system('minimus -D TGT=%s.afg %s.minimus' % (outroot, outroot))
if contig_prefix:
lines = open(outroot + '.minimus.contig').readlines()
fh = open(outroot + '.minimus.contig', 'w')
for l in lines:
if l.startswith('##'):
print >> fh, '##' + contig_prefix + l[2:],
else:
print >> fh, l,
fh.close()
if restore_singletons:
in_assem = re.findall('#(.+?)\(',
open(outroot + '.minimus.contig').read())
reads = Seq.Fasta(fasta)
for f in in_assem:
del reads[f]
all_fasta = outroot + '.all.fasta'
assem = Seq.Fasta(outroot + '.minimus.fasta')
allseq = Seq.Fasta()
allseq.update(dict([(contig_prefix + k, v) for k, v in assem.items()]))
allseq.update(reads)
allseq.write_to_file(all_fasta)
return all_fasta
else:
if contig_prefix:
f = outroot + '.minimus.fasta'
lines = open(f).readlines()
fh = open(f, 'w')
for l in lines:
if l.startswith('>'):
print >> fh, '>' + contig_prefix + l[1:],
else:
print >> fh, l,
fh.close()
return outroot + '.minimus.fasta'
def build_fasta_from_scaff_gff(infasta_s,
gff,
contig_prefix='',
include_singletons=True,
ol_minID=0.9,
outfile=None,
mum_len='4'):
'''takes scaffolding information from gff of the form generated by get_scaff_from_minimus
builds a single assembly for all scaffold instructions pertaining to seqids in infasta
if include_singletons is True, adds all sequences from infasta not included in scaffolds along with the scaffolded sequence in the returned assembly
ol_minID is the minimum %ID accepted for overlaps in contigs
'''
if isinstance(infasta_s, str):
infasta = Seq.Fasta(infasta_s)
else:
infasta = deepcopy(infasta_s)
suffixes = ['', 'b', 'c', 'd', 'e', 'f']
current_suffix = ''
#use only scaffolding info relevant to the specified infasta
in_ids = infasta.seq_names()
this_gff = [r for r in gff if r['seqid'] in in_ids]
#use only scaffolding info that joins 2 or more seqs
contigs = {}.fromkeys([r['attribute_contig'] for r in this_gff], 0)
for r in this_gff:
contigs[r['attribute_contig']] += 1
#get final ordered scaffolding layout
this_gff = sorted(
[r for r in this_gff if contigs[r['attribute_contig']] > 1],
key=lambda r: (r['attribute_contig'], int(r['attribute_cstart']),
int(r['attribute_cend'])))
#extract sequences and orient for scaffolding
assem_frags = infasta.substr_from_gff(this_gff,
plus_strand=True,
name_key=None)
assem = Seq.Fasta()
for k, v in contigs.items():
if v > 1:
assem[contig_prefix + k] = Seq.Sequence('')
if this_gff:
for i, r in enumerate(this_gff[:-1]):
next = this_gff[i + 1]
if r['attribute_contig'] == next['attribute_contig']:
s1, e1, s2, e2 = [
int(n) for n in [
r['attribute_cstart'], r['attribute_cend'],
next['attribute_cstart'], next['attribute_cend']
]
]
print >> sys.stderr, 'scaffolding %s %s %s %s:\n\t%s\n\t%s' % (
s1, e1, s2, e2, r, next)
if e1 > s2:
print >> sys.stderr, 'OVERLAP:\n\t%s\n\t%s' % (
assem_frags[r['seqid']][(s2 - s1):],
assem_frags[next['seqid']][:(e1 - s2 + 1)])
fa1 = Seq.Fasta()
fa2 = Seq.Fasta()
fa1['seq1'] = assem_frags[r['seqid']][(s2 - s1):]
fa2['seq2'] = assem_frags[next['seqid']][:(e1 - s2 + 1)]
shorter = min(len(fa1['seq1']), len(fa2['seq2']))
mums = Aln.mum(fa1,
fa2,
mumargs={'-l': '%s' % int(mum_len)})[0]
match = float(sum([mumr['score'] for mumr in mums]))
if (shorter <= 2*int(mum_len) + math.ceil((1-ol_minID)*shorter)) or \
(match/shorter >= ol_minID) or \
(fa1['seq1'][:shorter] == fa2['seq2'][:shorter]) or \
(Seq.is_simple(fa1['seq1']) or Seq.is_simple(fa2['seq2'])):
assem[contig_prefix + r['attribute_contig'] +
current_suffix] += assem_frags[r['seqid']][:s2]
else:
#implement record of splitting into a/b/etc fragments!
print >> sys.stdout, fa1, '\n', fa2, '\n', mums
current_suffix = suffixes[
suffixes.index(current_suffix) + 1]
print >> sys.stderr, 'overlap of %s bp %0.2f %%id unresolved (min %0.2f)\nstarting %s' % (
e1 - s2, match /
(e1 - s2), ol_minID, current_suffix)
assem[contig_prefix + r['attribute_contig'] +
current_suffix] = assem_frags[r['seqid']]
else:
spacer = Seq.Sequence('n' * (s2 - e1))
assem[contig_prefix + r['attribute_contig'] +
current_suffix] += assem_frags[r['seqid']] + spacer
else:
assem[contig_prefix + r['attribute_contig'] +
current_suffix] += assem_frags[r['seqid']]
current_suffix = ''
assem[contig_prefix + this_gff[-1]['attribute_contig'] +
current_suffix] += assem_frags[this_gff[-1]['seqid']]
if include_singletons:
singletons = dict([(k, v) for k, v in infasta.items()
if not k in [r['seqid'] for r in this_gff]])
assem.update(singletons)
if outfile:
assem.write_to_file(outfile)
return assem
