def test_no_reads():
merged_reads = prepare_reads()
builder = ConsensusBuilder()
list(builder.build(merged_reads))
with pytest.raises(IndexError):
builder.get_consensus()
def test_build_by_lengths_no_reads():
merged_reads = prepare_reads()
expected_consensuses = []
builder = ConsensusBuilder()
list(builder.build(merged_reads))
consensuses = list(builder.get_consensus_by_lengths())
assert expected_consensuses == consensuses
def test_build_by_lengths_with_no_neighbours():
raw_reads = ['A' * 120] * 9 + ['G' * 220] * 10
merged_reads = prepare_reads(*raw_reads)
expected_consensuses = ['G' * 220]
builder = ConsensusBuilder()
list(builder.build(merged_reads))
consensuses = list(builder.get_consensus_by_lengths())
assert expected_consensuses == consensuses
def test_different_lengths():
merged_reads = prepare_reads("AAACCCTTTGGGAAACCC", "ATACCCTTTGGGAAACCC",
"AAACCCTTTGGGAAACCC", "CATGAGACATCACACAC",
"CATGAGACATCACACA", "CATGAGACATCACAC",
"CATGAGACATCACA")
expected_consensus = "AAACCCTTTGGGAAACCC"
builder = ConsensusBuilder()
list(builder.build(merged_reads))
assert expected_consensus == builder.get_consensus()
def test_mixture():
merged_reads = list(
prepare_reads("AAACCCTTTGGGAAACCC", "ATACCCTATGGGAAACCC",
"AAACCCTTTGGGAAACCC"))
expected_consensus = "AAACCCTTTGGGAAACCC"
builder = ConsensusBuilder()
returned_reads = list(builder.build(merged_reads))
assert merged_reads == returned_reads
assert expected_consensus == builder.get_consensus()
def test_tie():
""" If there's an exact 50/50 mixture, pick alphabetically first. """
merged_reads = list(
prepare_reads("AAACCCTTTGGGAAACCC", "ATACCCTATGGGAAACCC",
"ATACCCTATGGGAAACCC", "AAACCCTTTGGGAAACCC"))
expected_consensus = "AAACCCTATGGGAAACCC"
builder = ConsensusBuilder()
returned_reads = list(builder.build(merged_reads))
assert merged_reads == returned_reads
assert expected_consensus == builder.get_consensus()
def test_build_by_lengths_when_spike_too_short():
raw_reads = []
for n in range(100, 141):
if n == 120:
raw_reads += ['A' * n] * 99
else:
raw_reads += ['C' * n] * 5
merged_reads = prepare_reads(*raw_reads)
expected_consensuses = []
builder = ConsensusBuilder()
list(builder.build(merged_reads))
consensuses = list(builder.get_consensus_by_lengths())
assert expected_consensuses == consensuses
def fastq_g2p(pssm,
fastq1,
fastq2,
g2p_csv,
g2p_summary_csv=None,
unmapped1=None,
unmapped2=None,
aligned_csv=None,
min_count=1,
min_valid=1,
min_valid_percent=0.0,
merged_contigs_csv=None):
g2p_filename = getattr(g2p_csv, 'name', None)
if g2p_filename is None:
count_prefix = None
else:
working_path = os.path.dirname(g2p_csv.name)
count_prefix = os.path.join(working_path, 'read_counts')
project_config = ProjectConfig.loadDefault()
hiv_seed = project_config.getReference(G2P_SEED_NAME)
coordinate_ref = project_config.getReference(COORDINATE_REF_NAME)
v3loop_ref = extract_target(hiv_seed, coordinate_ref)
reader = FastqReader(fastq1, fastq2)
merged_reads = merge_reads(reader)
consensus_builder = ConsensusBuilder()
counted_reads = consensus_builder.build(merged_reads)
trimmed_reads = trim_reads(counted_reads, v3loop_ref)
mapped_reads = write_unmapped_reads(trimmed_reads, unmapped1, unmapped2)
read_counts = count_reads(mapped_reads, count_prefix)
if aligned_csv is not None:
read_counts = write_aligned_reads(read_counts, aligned_csv, hiv_seed,
v3loop_ref)
write_rows(pssm,
read_counts,
g2p_csv,
g2p_summary_csv,
min_count,
min_valid=min_valid,
min_valid_percent=min_valid_percent)
if merged_contigs_csv is not None:
contig_writer = DictWriter(merged_contigs_csv, ['contig'])
contig_writer.writeheader()
for consensus in consensus_builder.get_consensus_by_lengths():
unambiguous_consensus = consensus.replace('N', '').replace('-', '')
if unambiguous_consensus:
contig_writer.writerow(dict(contig=consensus))
def test_build_by_lengths():
raw_reads = []
for n in range(100, 141):
if n == 120:
raw_reads += ['A' * n] * 100
else:
raw_reads += ['C' * n] * 2
for n in range(200, 241):
if n == 220:
raw_reads += ['G' * n] * 1000
else:
raw_reads += ['T' * n] * 20
merged_reads = prepare_reads(*raw_reads)
expected_consensuses = ['A' * 120, 'G' * 220]
builder = ConsensusBuilder()
list(builder.build(merged_reads))
consensuses = list(builder.get_consensus_by_lengths())
assert expected_consensuses == consensuses