def test_somatic_variant_with_2_supporting_rna_reads():
variant = Variant("14", 105849746, "G", "A")
base_dir = "data/somatic-variant-with-2-supporting-rna-reads/"
normal_reads = load_bam(base_dir +
"normal.14.105849746.G.A.no-alt.sorted.bam")
tumor_reads = load_bam(base_dir +
"tumor.14.105849746.G.A.many-alt.sorted.bam")
rna_reads = load_bam(base_dir + "rna.14.105849746.G.A.2-alt.sorted.bam")
normal_sample_variant_reads = reads_supporting_variant(
variant=variant, samfile=normal_reads)
eq_(len(normal_sample_variant_reads), 0)
print(normal_sample_variant_reads)
tumor_sample_variant_reads = reads_supporting_variant(variant=variant,
samfile=tumor_reads)
print(tumor_sample_variant_reads)
eq_(len(tumor_sample_variant_reads), 8)
rna_sample_variant_reads = reads_supporting_variant(variant=variant,
samfile=rna_reads)
print(rna_sample_variant_reads)
eq_(len(rna_sample_variant_reads), 2)
# Arun went through the hassle of pulling out the exact read names
# in IGV
expected_variant_rna_read_names = {
"K00193:50:H5NKVBBXX:5:2202:6421:24964",
"K00193:50:H5NKVBBXX:5:2119:30908:1138",
}
for variant_read in rna_sample_variant_reads:
assert variant_read.name in expected_variant_rna_read_names
def test_somatic_variant_with_2_supporting_rna_reads():
variant = Variant("14", 105849746, "G", "A")
base_dir = "data/somatic-variant-with-2-supporting-rna-reads/"
normal_reads = load_bam(base_dir + "normal.14.105849746.G.A.no-alt.sorted.bam")
tumor_reads = load_bam(base_dir + "tumor.14.105849746.G.A.many-alt.sorted.bam")
rna_reads = load_bam(base_dir + "rna.14.105849746.G.A.2-alt.sorted.bam")
normal_sample_variant_reads = reads_supporting_variant(
variant=variant,
samfile=normal_reads)
eq_(len(normal_sample_variant_reads), 0)
print(normal_sample_variant_reads)
tumor_sample_variant_reads = reads_supporting_variant(
variant=variant,
samfile=tumor_reads)
print(tumor_sample_variant_reads)
eq_(len(tumor_sample_variant_reads), 8)
rna_sample_variant_reads = reads_supporting_variant(
variant=variant,
samfile=rna_reads)
print(rna_sample_variant_reads)
eq_(len(rna_sample_variant_reads), 2)
# Arun went through the hassle of pulling out the exact read names
# in IGV
expected_variant_rna_read_names = {
"K00193:50:H5NKVBBXX:5:2202:6421:24964",
"K00193:50:H5NKVBBXX:5:2119:30908:1138",
}
for variant_read in rna_sample_variant_reads:
assert variant_read.name in expected_variant_rna_read_names
def test_assemble_transcript_fragments_snv():
samfile = load_bam("data/cancer-wgs-primary.chr12.bam")
chromosome = "chr12"
base1_location = 65857041
ref = "G"
alt = "C"
variant = Variant(
contig=chromosome,
start=base1_location,
ref=ref,
alt=alt,
ensembl=ensembl_grch38)
variant_reads = reads_supporting_variant(
variant=variant,
samfile=samfile,
chromosome=chromosome,)
sequences = iterative_overlap_assembly(
initial_variant_sequences_from_reads(variant_reads),
min_overlap_size=30)
assert len(sequences) > 0
max_read_length = max(len(r) for r in variant_reads)
for s in sequences:
print("%s%s%s weight=%d length=%d" % (
s.prefix,
s.alt,
s.suffix,
len(s.reads),
len(s.sequence)))
eq_(s.alt, alt)
assert len(s) > max_read_length, \
"Expected assembled sequences to be longer than read length (%d)" % (
max_read_length,)
def test_group_unique_sequences():
samfile = load_bam("data/cancer-wgs-primary.chr12.bam")
chromosome = "chr12"
base1_location = 65857041
ref = "G"
alt = "C"
variant = Variant(
contig=chromosome,
start=base1_location,
ref=ref, alt=alt,
ensembl=ensembl_grch38)
variant_reads = reads_supporting_variant(
samfile=samfile,
chromosome=chromosome,
variant=variant)
print("%d variant reads: %s" % (
len(variant_reads), variant_reads))
groups = group_unique_sequences(
variant_reads,
max_prefix_size=30,
max_suffix_size=30)
print("%d unique sequences: %s" % (
len(groups), groups))
# there are some redundant reads, so we expect that the number of
# unique entries should be less than the total read partitions
assert len(variant_reads) > len(groups)
def test_most_common_nucleotides_for_chr12_deletion():
samfile = load_bam("data/cancer-wgs-primary.chr12.bam")
chromosome = "chr12"
base1_location = 70091490
ref = "TTGTAGATGCTGCCTCTCC"
alt = ""
variant = Variant(
chromosome,
base1_location,
ref,
alt,
ensembl=ensembl_grch38)
variant_reads = reads_supporting_variant(
samfile=samfile,
chromosome=chromosome,
variant=variant)
consensus_sequence, chosen_counts, other_counts = most_common_nucleotides(
variant_reads)
print(chosen_counts)
print(other_counts)
eq_(len(chosen_counts), len(consensus_sequence))
eq_(len(other_counts), len(consensus_sequence))
assert other_counts.sum() < chosen_counts.sum(), \
"Counts for alternate nucleotides should not exceed the chosen sequence"
number_matching_reads = 0
for variant_read in variant_reads:
full_seq = variant_read.prefix + variant_read.allele + variant_read.suffix
number_matching_reads += (full_seq in consensus_sequence)
fraction_matching_reads = number_matching_reads / float(len(variant_reads))
print("Fraction matching reads is %d/%d = %f" % (
number_matching_reads, len(variant_reads), fraction_matching_reads))
assert fraction_matching_reads > 0.5, \
"Expected majority of reads to match consensus sequence"
def test_sequence_counts_snv():
samfile = load_bam("data/cancer-wgs-primary.chr12.bam")
chromosome = "chr12"
base1_location = 65857041
ref = "G"
alt = "C"
variant = Variant(chromosome, base1_location, ref, alt)
variant_reads = reads_supporting_variant(
samfile=samfile,
chromosome=chromosome,
variant=variant)
variant_sequences = reads_to_variant_sequences(
variant=variant,
reads=variant_reads,
preferred_sequence_length=61)
assert len(variant_sequences) == 1
for variant_sequence in variant_sequences:
print(variant_sequence)
eq_(variant_sequence.alt, alt)
eq_(len(variant_sequence.prefix), 30)
eq_(len(variant_sequence.suffix), 30)
eq_(
variant_sequence.prefix + variant_sequence.alt + variant_sequence.suffix,
variant_sequence.sequence)
def test_partitioned_read_sequences_deletion():
"""
test_partitioned_read_sequences_deletion : Test that read gets correctly
partitioned for chr1:4 TT>T where the sequence for chr1 is assumed to
be "ACCTTG"
"""
# chr1_seq = "ACCTTG"
chromosome = "chromosome"
location = 4
ref = "TT"
alt = "T"
variant = Variant(chromosome,
location,
ref,
alt,
normalize_contig_name=False)
read = make_read(seq="ACCTG", cigar="4M1D1M", mdtag="4^T1")
samfile = DummySamFile(reads=[read])
variant_reads = reads_supporting_variant(samfile=samfile,
chromosome=chromosome,
variant=variant)
print(variant_reads)
assert len(variant_reads) == 1
variant_read = variant_reads[0]
expected = AlleleRead(name=read.qname,
prefix="ACCT",
allele="",
suffix="G")
eq_(variant_read, expected)
def test_partitioned_read_sequences_deletion():
"""
test_partitioned_read_sequences_deletion : Test that read gets correctly
partitioned for chr1:4 TT>T where the sequence for chr1 is assumed to
be "ACCTTG"
"""
# chr1_seq = "ACCTTG"
chromosome = "chromosome"
location = 4
ref = "TT"
alt = "T"
variant = Variant(
chromosome, location, ref, alt, normalize_contig_name=False)
read = make_read(seq="ACCTG", cigar="4M1D1M", mdtag="4^T1")
samfile = DummySamFile(reads=[read])
variant_reads = reads_supporting_variant(
samfile=samfile,
chromosome=chromosome,
variant=variant)
print(variant_reads)
assert len(variant_reads) == 1
variant_read = variant_reads[0]
expected = AlleleRead(
name=read.qname,
prefix="ACCT",
allele="",
suffix="G")
eq_(variant_read, expected)
def test_assemble_transcript_fragments_snv():
samfile = load_bam("data/cancer-wgs-primary.chr12.bam")
chromosome = "chr12"
base1_location = 65857041
ref = "G"
alt = "C"
variant = Variant(contig=chromosome,
start=base1_location,
ref=ref,
alt=alt,
ensembl=ensembl_grch38)
variant_reads = reads_supporting_variant(
variant=variant,
samfile=samfile,
chromosome=chromosome,
)
sequences = iterative_overlap_assembly(
initial_variant_sequences_from_reads(variant_reads),
min_overlap_size=30)
assert len(sequences) > 0
max_read_length = max(len(r) for r in variant_reads)
for s in sequences:
print("%s%s%s weight=%d length=%d" %
(s.prefix, s.alt, s.suffix, len(s.reads), len(s.sequence)))
eq_(s.alt, alt)
if len(s.read_names) > 1:
# expect sequences supported by more than one read to be greater
# than the read length
assert len(s) > max_read_length, \
"Expected assembled sequences to be longer than read length (%d)" % (
max_read_length,)
def test_partition_variant_reads_deletion():
samfile = load_bam("data/cancer-wgs-primary.chr12.bam")
chromosome = "chr12"
base1_location = 70091490
ref = "TTGTAGATGCTGCCTCTCC"
alt = ""
variant = Variant(contig=chromosome, start=base1_location, ref=ref, alt=alt, ensembl=ensembl_grch38)
variant_reads = reads_supporting_variant(samfile=samfile, chromosome=chromosome, variant=variant)
assert len(variant_reads) > 1
for variant_read in variant_reads:
eq_(variant_read.allele, alt)
def test_somatic_variant_with_0_supporting_rna_reads():
variant = Variant("6", 90411765, "G", "A")
base_dir = "data/somatic-variant-with-0-supporting-rna-reads/"
normal_reads = load_bam(base_dir + "normal.6.90411765.G.A.sorted.bam")
tumor_reads = load_bam(base_dir + "tumor.6.90411765.G.A.sorted.bam")
rna_reads = load_bam(base_dir + "rna.6.90411765.G.A.sorted.bam")
normal_sample_variant_reads = reads_supporting_variant(
variant=variant, samfile=normal_reads)
eq_(len(normal_sample_variant_reads), 0)
print(normal_sample_variant_reads)
tumor_sample_variant_reads = reads_supporting_variant(variant=variant,
samfile=tumor_reads)
print(tumor_sample_variant_reads)
eq_(len(tumor_sample_variant_reads), 5)
rna_sample_variant_reads = reads_supporting_variant(variant=variant,
samfile=rna_reads)
print(rna_sample_variant_reads)
eq_(len(rna_sample_variant_reads), 0)
def test_partition_variant_reads_deletion():
samfile = load_bam("data/cancer-wgs-primary.chr12.bam")
chromosome = "chr12"
base1_location = 70091490
ref = "TTGTAGATGCTGCCTCTCC"
alt = ""
variant = Variant(
contig=chromosome,
start=base1_location,
ref=ref,
alt=alt,
ensembl=ensembl_grch38)
variant_reads = reads_supporting_variant(
samfile=samfile,
chromosome=chromosome,
variant=variant)
assert len(variant_reads) > 1
for variant_read in variant_reads:
eq_(variant_read.allele, alt)
def test_sequence_counts_snv():
samfile = load_bam("data/cancer-wgs-primary.chr12.bam")
chromosome = "chr12"
base1_location = 65857041
ref = "G"
alt = "C"
variant = Variant(chromosome, base1_location, ref, alt)
variant_reads = reads_supporting_variant(samfile=samfile,
chromosome=chromosome,
variant=variant)
variant_sequences = reads_to_variant_sequences(
variant=variant, reads=variant_reads, preferred_sequence_length=61)
assert len(variant_sequences) == 1
for variant_sequence in variant_sequences:
print(variant_sequence)
eq_(variant_sequence.alt, alt)
eq_(len(variant_sequence.prefix), 30)
eq_(len(variant_sequence.suffix), 30)
eq_(
variant_sequence.prefix + variant_sequence.alt +
variant_sequence.suffix, variant_sequence.sequence)