def test_simple_seg_file_input(self):
"""Test that we can read in a seg file, do no annotation, and output as SIMPLE_TSV"""
inputFilename = "testdata/seg/Patient0.seg.txt"
output_filename = "out/test_simple_seg_file_input.tsv"
if os.path.exists(output_filename):
os.remove(output_filename)
ic = MafliteInputMutationCreator(inputFilename, 'configs/seg_file_input.config')
segs = ic.createMutations()
i = 1
for i,seg in enumerate(segs):
pass
self.assertTrue((i+1) == 27, "Found %d segments when there should have been 27." % (i+1))
ic = MafliteInputMutationCreator(inputFilename, 'configs/seg_file_input.config')
segs = ic.createMutations()
outputRenderer = SimpleOutputRenderer(output_filename, '')
outputRenderer.renderMutations(segs)
# Now check the output
output_reader = GenericTsvReader(output_filename)
required_cols = ["Sample", "Num_Probes", "Segment_Mean"]
headers = output_reader.getFieldNames()
for rcol in required_cols:
self.assertTrue(rcol in headers)
for line_dict in output_reader:
self.assertTrue(line_dict['start'] is not None)
self.assertTrue(line_dict['start'].strip() != "")
self.assertTrue(line_dict['end'] is not None)
self.assertTrue(line_dict['end'].strip() != "")
def testSampleNameSelectorWithMaf(self):
input = MafliteInputMutationCreator("testdata/maflite/tiny_maflite.maf.txt")
first_mut = next(input.createMutations())
s = SampleNameSelector(first_mut)
for mut in input.createMutations():
self.assertEqual("Patient0-Normal-Patient0-Tumor", s.getSampleName(mut))
self.assertEqual(s.getAnnotationSource(),"OUTPUT")
self.assertEqual(s.getOutputAnnotationName(), MutUtils.SAMPLE_NAME_ANNOTATION_NAME)
def testSampleNameSelectorWithMaf(self):
input = MafliteInputMutationCreator(
"testdata/maflite/tiny_maflite.maf.txt")
first_mut = next(input.createMutations())
s = SampleNameSelector(first_mut)
for mut in input.createMutations():
self.assertEqual("Patient0-Normal-Patient0-Tumor",
s.getSampleName(mut))
self.assertEqual(s.getAnnotationSource(), "OUTPUT")
self.assertEqual(s.getOutputAnnotationName(),
MutUtils.SAMPLE_NAME_ANNOTATION_NAME)
def testFullIndelVcf(self):
""" Perform test of a Indel maflite all the way through TCGA VCF creation
"""
outputFilename = "out/TCGAVCFTest.indel.vcf"
callStatsIn = MafliteInputMutationCreator("testdata/maflite/Patient0.indel.maf.txt")
vcfOR = TcgaVcfOutputRenderer(outputFilename)
datasources = self._createDatasourcesForTesting()
annotator = Annotator()
annotator.setInputCreator(callStatsIn)
annotator.setOutputRenderer(vcfOR)
annotator.setManualAnnotations(self._createManualAnnotations())
for ds in datasources:
annotator.addDatasource(ds)
annotator.annotate()
self.assertTrue(os.path.exists(outputFilename))
# Check that the deletions have position decremented by one from what is present in the maflite
# Checking that 1 36643701 in the maflite (a deletion) becomes 1 36643700 in the vcf, but that the others are
# the same.
maflite_ic = MafliteInputMutationCreator("testdata/maflite/Patient0.indel.maf.txt")
muts = maflite_ic.createMutations()
vcf_reader = vcf.Reader(open(outputFilename, 'r'))
vcf_pos = [int(rec.POS) for rec in vcf_reader]
for m in muts:
# If the variant is a deletion, then the vcf position should be the same as maflite minus one. Otherwise, the same.
is_variant_deletion = (m.alt_allele == "") or (m.alt_allele == "-") or (m.alt_allele == ".")
if is_variant_deletion:
self.assertTrue((int(m.start) - 1) in vcf_pos, "Deletion was not correct for " + m.chr + ":" + m.start)
else:
self.assertTrue(int(m.start) in vcf_pos, "Insertion was not correct for " + m.chr + ":" + m.start)
def testFullIndelVcf(self):
""" Perform test of a Indel maflite all the way through TCGA VCF creation
"""
outputFilename = "out/TCGAVCFTest.indel.vcf"
callStatsIn = MafliteInputMutationCreator("testdata/maflite/Patient0.indel.maf.txt")
vcfOR = TcgaVcfOutputRenderer(outputFilename)
datasources = self._createDatasourcesForTesting()
annotator = Annotator()
annotator.setInputCreator(callStatsIn)
annotator.setOutputRenderer(vcfOR)
annotator.setManualAnnotations(self._createManualAnnotations())
for ds in datasources:
annotator.addDatasource(ds)
annotator.annotate()
self.assertTrue(os.path.exists(outputFilename))
# Check that the deletions have position decremented by one from what is present in the maflite
# Checking that 1 36643701 in the maflite (a deletion) becomes 1 36643700 in the vcf, but that the others are
# the same.
maflite_ic = MafliteInputMutationCreator("testdata/maflite/Patient0.indel.maf.txt")
muts = maflite_ic.createMutations()
vcf_reader = vcf.Reader(open(outputFilename, 'r'))
vcf_pos = [int(rec.POS) for rec in vcf_reader]
for m in muts:
# If the variant is a deletion, then the vcf position should be the same as maflite minus one. Otherwise, the same.
is_variant_deletion = (m.alt_allele == "") or (m.alt_allele == "-") or (m.alt_allele == ".")
if is_variant_deletion:
self.assertTrue((int(m.start) - 1) in vcf_pos, "Deletion was not correct for " + m.chr + ":" + m.start)
else:
self.assertTrue(int(m.start) in vcf_pos, "Insertion was not correct for " + m.chr + ":" + m.start)
def testFullSnpVcf(self):
""" Perform test of a SNP call stats (maflite) all the way through TCGA VCF creation. Only checks that a file was created.
"""
outputFilename = "out/TCGAVCFTest.snp.vcf"
callStatsIn = MafliteInputMutationCreator(
"testdata/Test.call_stats.trim.txt")
vcfOR = TcgaVcfOutputRenderer(outputFilename)
datasources = self._createDatasourcesForTesting()
annotator = Annotator()
annotator.setInputCreator(callStatsIn)
annotator.setOutputRenderer(vcfOR)
annotator.setManualAnnotations(self._createManualAnnotations())
for ds in datasources:
annotator.addDatasource(ds)
annotator.annotate()
self.assertTrue(os.path.exists(outputFilename))
maflite_ic = MafliteInputMutationCreator(
"testdata/maflite/Patient0.indel.maf.txt")
muts = maflite_ic.createMutations()
vcf_reader = vcf.Reader(open(outputFilename, 'r'))
for i, m in enumerate(muts):
rec = vcf_reader.next()
qual = rec.QUAL
# All records should have QUAL with a value (i.e. NOT ".")
self.assertIsNotNone(qual)
def test_simple_seg_file_annotations(self):
"""Test that we can read in a seg file, do GENCODE annotation, and output as SIMPLE_TSV"""
inputFilename = "testdata/seg/Patient0.seg.txt"
output_filename = "out/test_simple_seg_file_annotations.tsv"
if os.path.exists(output_filename):
os.remove(output_filename)
ic = MafliteInputMutationCreator(inputFilename, None,
'configs/seg_file_input.config')
segs = ic.createMutations()
i = 1
for i, seg in enumerate(segs):
pass
self.assertTrue(
(i + 1) == 27,
"Found %d segments when there should have been 27." % (i + 1))
ic = MafliteInputMutationCreator(inputFilename, None,
'configs/seg_file_input.config')
segs = ic.createMutations()
gencode_ds = TestUtils._create_test_gencode_v19_ds(
"out/seg_file_gencode_ds")
annotator = Annotator()
segs_annotated = []
for seg in segs:
segs_annotated.append(gencode_ds.annotate_segment(seg))
outputRenderer = SimpleOutputRenderer(output_filename, '')
outputRenderer.renderMutations(segs_annotated.__iter__())
# Now check the output
output_reader = GenericTsvReader(output_filename)
required_cols = ["Sample", "Num_Probes", "Segment_Mean"]
headers = output_reader.getFieldNames()
for rcol in required_cols:
self.assertTrue(rcol in headers)
for line_dict in output_reader:
self.assertTrue(line_dict['start'] is not None)
self.assertTrue(line_dict['start'].strip() != "")
self.assertTrue(line_dict['end'] is not None)
self.assertTrue(line_dict['end'].strip() != "")
self.assertTrue("genes" in line_dict.keys())
def test_alt1_vs_alt2(self):
"""Test that we pick up the alternate that is different from the reference when both are specified"""
ic = MafliteInputMutationCreator("testdata/maflite/alt1_vs_alt2.maflite")
muts = ic.createMutations()
ctr = 0
for m in muts:
ctr += 1
self.assertTrue(m.alt_allele == "C", "Did not properly populate the alternate allele in line " + str(ctr) + " " + m.alt_allele)
def testChromosomeM(self):
""" Make sure that the chromosome created as M, rather than MT."""
tmp = MafliteInputMutationCreator("testdata/maflite/chrM.maf.txt")
muts = tmp.createMutations()
for m in muts:
self.assertTrue(
m.chr == "M",
"mitochondria chromosome should be M, not " + m.chr)
def testSimpleRead(self):
""" Read a good maflite file and make sure that each mutation validates """
tmp = MafliteInputMutationCreator("testdata/maflite/Patient0.indel.maf.txt", 'configs/maflite_input.config')
muts = tmp.createMutations()
# If no exception is thrown, then this test passes.
for m in muts:
MutUtils.validateMutation(m)
def testTCGAMAFAsInput(self):
""" Test that we can take in a TCGA MAF (using MAFLITE), do no annotations, and still render it properly """
tmp = MafliteInputMutationCreator("testdata/maf/Patient0.maf.annotated", 'configs/maflite_input.config')
muts = tmp.createMutations()
outputFilename = "out/testTCGAMAFAsInput.tsv"
outputRenderer = TcgaMafOutputRenderer(outputFilename, 'configs/tcgaMAF2.4_output.config')
outputRenderer.renderMutations(muts, tmp.getComments())
def testNoUnknownAnnotations(self):
""" Make sure that the gaf 3.0 datasource does not annotate anything with source set to Unknown """
inputCreator = MafliteInputMutationCreator('testdata/maflite/Patient0.snp.maf.txt')
gafDatasource = TestUtils.createTranscriptProviderDatasource(self.config)
mutations = inputCreator.createMutations()
for m in mutations:
m = gafDatasource.annotate_mutation(m)
MutUtils.validateMutation(m)
unknownAnnotations = MutUtils.getUnknownAnnotations(m)
self.assertTrue(len(unknownAnnotations) == 0, "Unknown annotations exist in mutation: " + str(unknownAnnotations))
def testNumberOfMuts(self):
""" Make sure that the proper number of mutations were generated """
inputFilename = "testdata/maflite/Patient0.snp.maf.txt"
tmp = MafliteInputMutationCreator(inputFilename, 'configs/maflite_input.config')
muts = tmp.createMutations()
numMutsInput = len(file(inputFilename,'r').readlines()) - 1
ctr = 0
for m in muts:
ctr += 1
self.assertEqual(ctr, numMutsInput, "Did not see the proper number of mutations.")
def testSimpleRead(self):
""" Read a good maflite file and make sure that each mutation validates """
tmp = MafliteInputMutationCreator(
"testdata/maflite/Patient0.indel.maf.txt", None,
'configs/maflite_input.config')
muts = tmp.createMutations()
# If no exception is thrown, then this test passes.
for m in muts:
MutUtils.validateMutation(m)
def testNumberOfMuts(self):
""" Make sure that the proper number of mutations were generated """
inputFilename = "testdata/maflite/Patient0.snp.maf.txt"
tmp = MafliteInputMutationCreator(inputFilename)
muts = tmp.createMutations()
numMutsInput = len(file(inputFilename, 'r').readlines()) - 1
ctr = 0
for m in muts:
ctr += 1
self.assertEqual(ctr, numMutsInput,
"Did not see the proper number of mutations.")
def testTCGAMAFAsInput(self):
""" Test that we can take in a TCGA MAF (using MAFLITE), do no annotations, and still render it properly """
tmp = MafliteInputMutationCreator(
"testdata/maf/Patient0.maf.annotated", None,
'configs/maflite_input.config')
muts = tmp.createMutations()
outputFilename = "out/testTCGAMAFAsInput.tsv"
outputRenderer = TcgaMafOutputRenderer(
outputFilename, 'configs/tcgaMAF2.4_output.config')
outputRenderer.renderMutations(muts, tmp.getComments())
def test_alt1_vs_alt2(self):
"""Test that we pick up the alternate that is different from the reference when both are specified"""
ic = MafliteInputMutationCreator(
"testdata/maflite/alt1_vs_alt2.maflite")
muts = ic.createMutations()
ctr = 0
for m in muts:
ctr += 1
self.assertTrue(
m.alt_allele == "C",
"Did not properly populate the alternate allele in line " +
str(ctr) + " " + m.alt_allele)
def testChrGLs(self):
""" Test that mutations on unaligned transcripts can be annotated properly. I.e. when chromosome = GL....."""
inputCreator = MafliteInputMutationCreator('testdata/maflite/chrGLs.maf.tsv', "configs/maflite_input.config")
gafDatasource = TestUtils.createTranscriptProviderDatasource(self.config)
mutations = inputCreator.createMutations()
for m in mutations:
try:
m = gafDatasource.annotate_mutation(m)
MutUtils.validateMutation(m)
except Exception as e:
# Fail this test because an exception was thrown
self.assertTrue(False, "Erroneous exception was thrown: " + str(e) + "\n" + traceback.format_exc())
self.assertTrue(m['gene'] != '')
def testNoLostMutations(self):
""" Does a simple gaf datasource annotation run and makes sure that no mutations were lost """
inputFilename = 'testdata/maflite/Patient0.snp.maf.txt'
inputCreator = MafliteInputMutationCreator(inputFilename, "configs/maflite_input.config")
gafDatasource = TestUtils.createTranscriptProviderDatasource(self.config)
numMutsInput = len(file(inputFilename, 'r').readlines()) - 1
mutations = inputCreator.createMutations()
ctr = 0
for m in mutations:
m = gafDatasource.annotate_mutation(m)
MutUtils.validateMutation(m)
ctr += 1
self.assertEqual(ctr, numMutsInput, "Gaf data source altered mutation count.")
def testNoUnknownAnnotations(self):
""" Make sure that the gaf 3.0 datasource does not annotate anything with source set to Unknown """
inputCreator = MafliteInputMutationCreator(
'testdata/maflite/Patient0.snp.maf.txt')
gafDatasource = TestUtils.createTranscriptProviderDatasource(
self.config)
mutations = inputCreator.createMutations()
for m in mutations:
m = gafDatasource.annotate_mutation(m)
MutUtils.validateMutation(m)
unknownAnnotations = MutUtils.getUnknownAnnotations(m)
self.assertTrue(
len(unknownAnnotations) == 0,
"Unknown annotations exist in mutation: " +
str(unknownAnnotations))
def testMulticoreAnnotateFromChunkedFile(self):
#TODO: Add unit test that Mutation data is pickle-able
inputFile = "testdata/maflite/Patient0.snp.maf.txt"
outputFile = "out/testGAFMulticorePatient0.snp.maf.txt"
chunkSize = 200
numChunks = 4
gafDatasource = TestUtils.createGafDatasourceProxy(self.config)
ic = MafliteInputMutationCreator(inputFile)
oc = SimpleOutputRenderer(outputFile)
# createChunks
muts = ic.createMutations()
allAnnotatedChunksFlat = []
are_mutations_remaining = True
p = LoggingPool(processes=numChunks)
while are_mutations_remaining:
chunks = []
for j in xrange(0, numChunks):
chunk = []
for i in xrange(0, chunkSize):
try:
chunk.append(muts.next())
except StopIteration:
are_mutations_remaining = False
break
chunks.append((chunk, gafDatasource))
annotatedChunks = p.map(annotate_mutations_global, chunks)
annotatedChunksFlat = self._flattenChunks(annotatedChunks)
allAnnotatedChunksFlat.append(annotatedChunksFlat)
p.close()
p.join()
annotatedMuts = chain.from_iterable(allAnnotatedChunksFlat)
ctr = 0
oc.renderMutations(annotatedMuts, Metadata())
tsvReader = GenericTsvReader(outputFile)
for line in tsvReader:
ctr += 1
self.assertTrue(ctr == 730,
"Should have read 730 variants, but read " + str(ctr))
def testNoLostMutations(self):
""" Does a simple gaf datasource annotation run and makes sure that no mutations were lost """
inputFilename = 'testdata/maflite/Patient0.snp.maf.txt'
inputCreator = MafliteInputMutationCreator(
inputFilename, "configs/maflite_input.config")
gafDatasource = TestUtils.createTranscriptProviderDatasource(
self.config)
numMutsInput = len(file(inputFilename, 'r').readlines()) - 1
mutations = inputCreator.createMutations()
ctr = 0
for m in mutations:
m = gafDatasource.annotate_mutation(m)
MutUtils.validateMutation(m)
ctr += 1
self.assertEqual(ctr, numMutsInput,
"Gaf data source altered mutation count.")
def testChrGLs(self):
""" Test that mutations on unaligned transcripts can be annotated properly. I.e. when chromosome = GL....."""
inputCreator = MafliteInputMutationCreator(
'testdata/maflite/chrGLs.maf.tsv', "configs/maflite_input.config")
gafDatasource = TestUtils.createTranscriptProviderDatasource(
self.config)
mutations = inputCreator.createMutations()
for m in mutations:
try:
m = gafDatasource.annotate_mutation(m)
MutUtils.validateMutation(m)
except Exception as e:
# Fail this test because an exception was thrown
self.assertTrue(
False, "Erroneous exception was thrown: " + str(e) + "\n" +
traceback.format_exc())
self.assertTrue(m['gene'] != '')
def testMulticoreAnnotateFromChunkedFile(self):
#TODO: Add unit test that Mutation data is pickle-able
inputFile = "testdata/maflite/Patient0.snp.maf.txt"
outputFile = "out/testGAFMulticorePatient0.snp.maf.txt"
chunkSize = 200
numChunks = 4
gafDatasource = TestUtils.createGafDatasourceProxy(self.config)
ic = MafliteInputMutationCreator(inputFile)
oc = SimpleOutputRenderer(outputFile)
# createChunks
muts = ic.createMutations()
allAnnotatedChunksFlat = []
are_mutations_remaining = True
p = LoggingPool(processes=numChunks)
while are_mutations_remaining:
chunks = []
for j in xrange(0, numChunks):
chunk = []
for i in xrange(0, chunkSize):
try:
chunk.append(muts.next())
except StopIteration:
are_mutations_remaining = False
break
chunks.append((chunk, gafDatasource))
annotatedChunks = p.map(annotate_mutations_global, chunks)
annotatedChunksFlat = self._flattenChunks(annotatedChunks)
allAnnotatedChunksFlat.append(annotatedChunksFlat)
p.close()
p.join()
annotatedMuts = chain.from_iterable(allAnnotatedChunksFlat)
ctr = 0
oc.renderMutations(annotatedMuts, Metadata())
tsvReader = GenericTsvReader(outputFile)
for line in tsvReader:
ctr += 1
self.assertTrue(ctr == 730, "Should have read 730 variants, but read " + str(ctr))
def testChromosomeM(self):
""" Make sure that the chromosome created as M, rather than MT."""
tmp = MafliteInputMutationCreator("testdata/maflite/chrM.maf.txt", 'configs/maflite_input.config')
muts = tmp.createMutations()
for m in muts:
self.assertTrue(m.chr=="M", "mitochondria chromosome should be M, not " + m.chr)