def get_consensus_with_trim( c_input ):
seqs, seed_id, config = c_input
min_cov, K, local_match_count_window, local_match_count_threshold, max_n_read, min_idt, edge_tolerance, trim_size = config
trim_seqs = []
seed = seqs[0]
for seq in seqs[1:]:
aln_data = get_alignment(seq, seed, edge_tolerance)
s1, e1, s2, e2, aln_size, aln_score, c_status = aln_data
if c_status == "none":
continue
if aln_score > 1000 and e1 - s1 > 500:
e1 -= trim_size
s1 += trim_size
trim_seqs.append( (e1-s1, seq[s1:e1]) )
trim_seqs.sort(key = lambda x:-x[0]) #use longest alignment first
trim_seqs = [x[1] for x in trim_seqs]
if len(trim_seqs) > max_n_read:
trim_seqs = trim_seqs[:max_n_read]
trim_seqs = [seed] + trim_seqs
seqs_ptr = (c_char_p * len(trim_seqs))()
seqs_ptr[:] = trim_seqs
consensus_data_ptr = falcon.generate_consensus( seqs_ptr, len(trim_seqs), min_cov, K,
local_match_count_window, local_match_count_threshold, min_idt )
consensus = string_at(consensus_data_ptr[0].sequence)[:]
eff_cov = consensus_data_ptr[0].eff_cov[:len(consensus)]
falcon.free_consensus_data( consensus_data_ptr )
del seqs_ptr
return consensus, seed_id
def get_consensus_without_trim( c_input ):
seqs, seed_id, config = c_input
min_cov, K, max_n_read, min_idt, edge_tolerance, trim_size, min_cov_aln, max_cov_aln = config
if len(seqs) > max_n_read:
seqs = get_longest_reads(seqs, max_n_read, max_cov_aln, sort=True)
seqs_ptr = (c_char_p * len(seqs))()
seqs_ptr[:] = seqs
consensus_data_ptr = falcon.generate_consensus( seqs_ptr, len(seqs), min_cov, K, min_idt )
consensus = string_at(consensus_data_ptr[0].sequence)[:]
eff_cov = consensus_data_ptr[0].eff_cov[:len(consensus)]
falcon.free_consensus_data( consensus_data_ptr )
del seqs_ptr
return consensus, seed_id
def get_consensus_without_trim( c_input ):
seqs, seed_id, config = c_input
min_cov, K, local_match_count_window, local_match_count_threshold, max_n_read, min_idt, edge_tolerance, trim_size = config
if len(seqs) > max_n_read:
seqs = seqs[:max_n_read]
seqs_ptr = (c_char_p * len(seqs))()
seqs_ptr[:] = seqs
consensus_data_ptr = falcon.generate_consensus( seqs_ptr, len(seqs), min_cov, K,
local_match_count_window, local_match_count_threshold, min_idt )
consensus = string_at(consensus_data_ptr[0].sequence)[:]
eff_cov = consensus_data_ptr[0].eff_cov[:len(consensus)]
falcon.free_consensus_data( consensus_data_ptr )
del seqs_ptr
return consensus, seed_id
def get_consensus_without_trim(c_input):
seqs, seed_id, config = c_input
min_cov, K, max_n_read, min_idt, edge_tolerance, trim_size, min_cov_aln, max_cov_aln = config
if len(seqs) > max_n_read:
seqs = get_longest_reads(seqs, max_n_read, max_cov_aln, sort=True)
seqs_ptr = (c_char_p * len(seqs))()
seqs_ptr[:] = seqs
consensus_data_ptr = falcon.generate_consensus(seqs_ptr, len(seqs),
min_cov, K, min_idt)
consensus = string_at(consensus_data_ptr[0].sequence)[:]
eff_cov = consensus_data_ptr[0].eff_cov[:len(consensus)]
falcon.free_consensus_data(consensus_data_ptr)
del seqs_ptr
return consensus, seed_id
def get_consensus_without_trim(c_input):
seqs, seed_id, config = c_input
LOG.debug('Starting get_consensus_without_trim(len(seqs)=={}, seed_id={})'.format(
len(seqs), seed_id))
min_cov, K, max_n_read, min_idt, edge_tolerance, trim_size, min_cov_aln, max_cov_aln = config
if len(seqs) > max_n_read:
seqs = get_longest_reads(seqs, max_n_read, max_cov_aln, sort=True)
seqs_ptr = (c_char_p * len(seqs))()
seqs_ptr[:] = seqs
consensus_data_ptr = falcon.generate_consensus(
seqs_ptr, len(seqs), min_cov, K, min_idt)
assert consensus_data_ptr
consensus = string_at(consensus_data_ptr[0].sequence)[:]
eff_cov = consensus_data_ptr[0].eff_cov[:len(consensus)]
LOG.debug(' Freeing1')
falcon.free_consensus_data(consensus_data_ptr)
del seqs_ptr
LOG.debug(' Finishing get_consensus_without_trim(seed_id={})'.format(seed_id))
return consensus, seed_id
def get_consensus_without_trim(c_input):
seqs, seed_id, config = c_input
LOG.debug('Starting get_consensus_without_trim(len(seqs)=={}, seed_id={})'.
format(len(seqs), seed_id))
min_cov, K, max_n_read, min_idt, edge_tolerance, trim_size, min_cov_aln, max_cov_aln = config
if len(seqs) > max_n_read:
seqs = get_longest_reads(seqs, max_n_read, max_cov_aln, sort=True)
seqs_ptr = (c_char_p * len(seqs))()
seqs_ptr[:] = seqs
consensus_data_ptr = falcon.generate_consensus(seqs_ptr, len(seqs),
min_cov, K, min_idt)
consensus = string_at(consensus_data_ptr[0].sequence)[:]
eff_cov = consensus_data_ptr[0].eff_cov[:len(consensus)]
LOG.debug(' Freeing1')
falcon.free_consensus_data(consensus_data_ptr)
del seqs_ptr
LOG.debug(
' Finishing get_consensus_without_trim(seed_id={})'.format(seed_id))
return consensus, seed_id
def get_consensus_with_trim(c_input):
seqs, seed_id, config = c_input
LOG.debug(
'Starting get_consensus_with_trim(len(seqs)=={}, seed_id={})'.format(
len(seqs), seed_id))
min_cov, K, max_n_read, min_idt, edge_tolerance, trim_size, min_cov_aln, max_cov_aln = config
trim_seqs = []
seed = seqs[0]
for seq in seqs[1:]:
aln_data = get_alignment(seq, seed, edge_tolerance)
s1, e1, s2, e2, aln_size, aln_score, c_status = aln_data
if c_status == "none":
continue
if aln_score > 1000 and e1 - s1 > 500:
e1 -= trim_size
s1 += trim_size
trim_seqs.append((e1 - s1, seq[s1:e1]))
trim_seqs.sort(key=lambda x: -x[0]) # use longest alignment first
trim_seqs = [x[1] for x in trim_seqs]
trim_seqs = [seed] + trim_seqs
if len(trim_seqs[1:]) > max_n_read:
# seqs already sorted, dont' sort again
trim_seqs = get_longest_reads(trim_seqs,
max_n_read,
max_cov_aln,
sort=False)
seqs_ptr = (c_char_p * len(trim_seqs))()
seqs_ptr[:] = trim_seqs
consensus_data_ptr = falcon.generate_consensus(seqs_ptr, len(trim_seqs),
min_cov, K, min_idt)
consensus = string_at(consensus_data_ptr[0].sequence)[:]
eff_cov = consensus_data_ptr[0].eff_cov[:len(consensus)]
LOG.debug(' Freeing2')
falcon.free_consensus_data(consensus_data_ptr)
del seqs_ptr
LOG.debug(' Finishing get_consensus_with_trim(seed_id={})'.format(seed_id))
return consensus, seed_id
def get_consensus_core(seqs, min_cov, K, min_idt, allow_external_mapping):
seqs_ptr = (c_char_p * len(seqs))()
seqs_ptr[:] = [bytes(val.seq, encoding='ascii') for val in seqs]
all_seqs_mapped = False
if allow_external_mapping:
all_seqs_mapped = True
for seq in seqs:
if not seq.is_mapped:
all_seqs_mapped = False
break
if not all_seqs_mapped:
LOG.info('Internally mapping the sequences.')
consensus_data_ptr = falcon.generate_consensus(
seqs_ptr, len(seqs), min_cov, K, min_idt)
else:
LOG.info('Using external mapping coordinates from input.')
aln_ranges_ptr = (POINTER(falcon_kit.AlnRange) * len(seqs))()
for i, seq in enumerate(seqs):
a = falcon_kit.AlnRange(seq.qstart, seq.qend, seq.tstart, seq.tend, (seq.qend - seq.qstart))
aln_ranges_ptr[i] = pointer(a)
consensus_data_ptr = falcon.generate_consensus_from_mapping(
seqs_ptr, aln_ranges_ptr, len(seqs), min_cov, K, min_idt)
del aln_ranges_ptr
del seqs_ptr
if not consensus_data_ptr:
return ''
# assert consensus_data_ptr
consensus = string_at(consensus_data_ptr[0].sequence)[:]
#eff_cov = consensus_data_ptr[0].eff_cov[:len(consensus)]
LOG.debug(' Freeing')
falcon.free_consensus_data(consensus_data_ptr)
return consensus.decode('ascii')
def get_consensus_with_trim(c_input):
seqs, seed_id, config = c_input
LOG.debug('Starting get_consensus_with_trim(len(seqs)=={}, seed_id={})'.format(
len(seqs), seed_id))
min_cov, K, max_n_read, min_idt, edge_tolerance, trim_size, min_cov_aln, max_cov_aln = config
trim_seqs = []
seed = seqs[0]
for seq in seqs[1:]:
aln_data = get_alignment(seq, seed, edge_tolerance)
s1, e1, s2, e2, aln_size, aln_score, c_status = aln_data
if c_status == "none":
continue
if aln_score > 1000 and e1 - s1 > 500:
e1 -= trim_size
s1 += trim_size
trim_seqs.append((e1 - s1, seq[s1:e1]))
trim_seqs.sort(key=lambda x: -x[0]) # use longest alignment first
trim_seqs = [x[1] for x in trim_seqs]
trim_seqs = [seed] + trim_seqs
if len(trim_seqs[1:]) > max_n_read:
# seqs already sorted, dont' sort again
trim_seqs = get_longest_reads(
trim_seqs, max_n_read, max_cov_aln, sort=False)
seqs_ptr = (c_char_p * len(trim_seqs))()
seqs_ptr[:] = trim_seqs
consensus_data_ptr = falcon.generate_consensus(
seqs_ptr, len(trim_seqs), min_cov, K, min_idt)
assert consensus_data_ptr
consensus = string_at(consensus_data_ptr[0].sequence)[:]
eff_cov = consensus_data_ptr[0].eff_cov[:len(consensus)]
LOG.debug(' Freeing2')
falcon.free_consensus_data(consensus_data_ptr)
del seqs_ptr
LOG.debug(' Finishing get_consensus_with_trim(seed_id={})'.format(seed_id))
return consensus, seed_id
