def main(recordFilenames, fastaFilename, title, xRange, bitRange):
"""
Print reads that match in a specified X-axis and bit score range.
@param recordFilenames: A C{list} of C{str} file names contain results of a
BLAST run, in JSON format.
@param fastaFilename: The C{str} name of the FASTA file that was originally
BLASTed.
@param title: The C{str} title of the subject sequence, as output by BLAST.
@param xRange: A (start, end) list of C{int}s, giving an X-axis range or
C{None} if the entire X axis range should be printed.
@param bitRange: A (start, end) list of C{int}s, giving a bit score range
or C{None} if the entire bit score range should be printed.
"""
reads = FastaReads(fastaFilename)
blastReadsAlignments = BlastReadsAlignments(reads, recordFilenames)
filtered = blastReadsAlignments.filter(whitelist=set([title]),
negativeTitleRegex='.')
titlesAlignments = TitlesAlignments(filtered)
if title not in titlesAlignments:
print('%s: Title %r not found in BLAST output' % (sys.argv[0], title))
sys.exit(3)
for titleAlignment in titlesAlignments[title]:
for hsp in titleAlignment.hsps:
if ((xRange is None or (xRange[0] <= hsp.subjectEnd and
xRange[1] >= hsp.subjectStart)) and
(bitRange is None or (bitRange[0] <= hsp.score.score <=
bitRange[1]))):
print(('query: %s, start: %d, end: %d, score: %d' % (
titleAlignment.read.id, hsp.subjectStart,
hsp.subjectEnd, hsp.score.score)))
def main(recordFilenames, fastaFilename, title, xRange, bitRange):
"""
Print reads that match in a specified X-axis and bit score range.
@param recordFilenames: A C{list} of C{str} file names contain results of a
BLAST run, in JSON format.
@param fastaFilename: The C{str} name of the FASTA file that was originally
BLASTed.
@param title: The C{str} title of the subject sequence, as output by BLAST.
@param xRange: A (start, end) list of C{int}s, giving an X-axis range or
C{None} if the entire X axis range should be printed.
@param bitRange: A (start, end) list of C{int}s, giving a bit score range
or C{None} if the entire bit score range should be printed.
"""
reads = FastaReads(fastaFilename)
blastReadsAlignments = BlastReadsAlignments(reads, recordFilenames)
filtered = blastReadsAlignments.filter(whitelist=set([title]),
negativeTitleRegex='.')
titlesAlignments = TitlesAlignments(filtered)
if title not in titlesAlignments:
print('%s: Title %r not found in BLAST output' % (sys.argv[0], title))
sys.exit(3)
for titleAlignment in titlesAlignments[title]:
for hsp in titleAlignment.hsps:
if ((xRange is None or (xRange[0] <= hsp.subjectEnd
and xRange[1] >= hsp.subjectStart))
and (bitRange is None or
(bitRange[0] <= hsp.score.score <= bitRange[1]))):
print(('query: %s, start: %d, end: %d, score: %d' %
(titleAlignment.read.id, hsp.subjectStart,
hsp.subjectEnd, hsp.score.score)))
def testCoverageIncludesSome(self):
"""
The coverage function must return an titlesAlignments instance with
only the expected titles if only some of its titles have sufficient
coverage.
"""
mockOpener = mockOpen(
read_data=(dumps(PARAMS) + '\n' + dumps(RECORD0) + '\n' +
dumps(RECORD1) + '\n' + dumps(RECORD2) + '\n' +
dumps(RECORD3) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
reads.add(Read('id0', 'A' * 70))
reads.add(Read('id1', 'A' * 70))
reads.add(Read('id2', 'A' * 70))
reads.add(Read('id3', 'A' * 70))
readsAlignments = BlastReadsAlignments(reads, 'file.json')
titlesAlignments = TitlesAlignments(readsAlignments)
# To understand why the following produces the result it does,
# you need to look at the HSP coverage in sample_data.py and
# calculate the coverage by hand.
result = titlesAlignments.filter(minCoverage=0.0011)
self.assertEqual([
'gi|887699|gb|DQ37780 Cowpox virus 15',
'gi|887699|gb|DQ37780 Monkeypox virus 456',
'gi|887699|gb|DQ37780 Mummypox virus 3000 B.C.',
], sorted(result.keys()))
def testCoverageIncludesAll(self):
"""
The coverage function must return an titlesAlignments instance with
all titles if all its titles has sufficient coverage.
"""
mockOpener = mockOpen(
read_data=(dumps(PARAMS) + '\n' + dumps(RECORD0) + '\n' +
dumps(RECORD1) + '\n' + dumps(RECORD2) + '\n' +
dumps(RECORD3) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
reads.add(Read('id0', 'A' * 70))
reads.add(Read('id1', 'A' * 70))
reads.add(Read('id2', 'A' * 70))
reads.add(Read('id3', 'A' * 70))
readsAlignments = BlastReadsAlignments(reads, 'file.json')
titlesAlignments = TitlesAlignments(readsAlignments)
result = titlesAlignments.filter(minCoverage=0.0)
self.assertEqual([
'gi|887699|gb|DQ37780 Cowpox virus 15',
'gi|887699|gb|DQ37780 Monkeypox virus 456',
'gi|887699|gb|DQ37780 Mummypox virus 3000 B.C.',
'gi|887699|gb|DQ37780 Squirrelpox virus 1296/99',
'gi|887699|gb|DQ37780 Squirrelpox virus 55',
], sorted(result.keys()))
def testReadSetFilterAllowAnything(self):
"""
The filter function work correctly when passed a 0.0 value for
minNewReads, i.e. that considers any read set sufficiently novel.
"""
mockOpener = mockOpen(
read_data=(dumps(PARAMS) + '\n' + dumps(RECORD0) + '\n' +
dumps(RECORD1) + '\n' + dumps(RECORD2) + '\n' +
dumps(RECORD3) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
reads.add(Read('id0', 'A' * 70))
reads.add(Read('id1', 'A' * 70))
reads.add(Read('id2', 'A' * 70))
reads.add(Read('id3', 'A' * 70))
readsAlignments = BlastReadsAlignments(reads, 'file.json')
titlesAlignments = TitlesAlignments(readsAlignments)
result = titlesAlignments.filter(minNewReads=0.0)
self.assertEqual([
'gi|887699|gb|DQ37780 Cowpox virus 15',
'gi|887699|gb|DQ37780 Monkeypox virus 456',
'gi|887699|gb|DQ37780 Mummypox virus 3000 B.C.',
'gi|887699|gb|DQ37780 Squirrelpox virus 1296/99',
'gi|887699|gb|DQ37780 Squirrelpox virus 55',
], sorted(result.keys()))
def testMedianScore_Bits(self):
"""
Sorting on median score must work when scores are bit scores,
including a secondary sort on title.
"""
mockOpener = mockOpen(
read_data=(dumps(PARAMS) + '\n' + dumps(RECORD0) + '\n' +
dumps(RECORD1) + '\n' + dumps(RECORD2) + '\n' +
dumps(RECORD3) + '\n' + dumps(RECORD4) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
reads.add(Read('id0', 'A' * 70))
reads.add(Read('id1', 'A' * 70))
reads.add(Read('id2', 'A' * 70))
reads.add(Read('id3', 'A' * 70))
reads.add(Read('id4', 'A' * 70))
readsAlignments = BlastReadsAlignments(reads, 'file.json')
titlesAlignments = TitlesAlignments(readsAlignments)
result = titlesAlignments.sortTitles('medianScore')
self.assertEqual(
[
'gi|887699|gb|DQ37780 Squirrelpox virus 55', # 25
'gi|887699|gb|DQ37780 Monkeypox virus 456', # 20
'gi|887699|gb|DQ37780 Mummypox virus 3000 B.C.', # 20
'gi|887699|gb|DQ37780 Squirrelpox virus 1296/99', # 20
'gi|887699|gb|DQ37780 Cowpox virus 15', # 20
],
result)
def testTitle(self):
"""
Sorting on title must work.
"""
mockOpener = mockOpen(
read_data=(dumps(PARAMS) + '\n' + dumps(RECORD0) + '\n' +
dumps(RECORD1) + '\n' + dumps(RECORD2) + '\n' +
dumps(RECORD3) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
reads.add(Read('id0', 'A' * 70))
reads.add(Read('id1', 'A' * 70))
reads.add(Read('id2', 'A' * 70))
reads.add(Read('id3', 'A' * 70))
readsAlignments = BlastReadsAlignments(
reads, 'file.json', scoreClass=LowerIsBetterScore)
titlesAlignments = TitlesAlignments(readsAlignments)
result = titlesAlignments.sortTitles('title')
self.assertEqual([
'gi|887699|gb|DQ37780 Cowpox virus 15',
'gi|887699|gb|DQ37780 Monkeypox virus 456',
'gi|887699|gb|DQ37780 Mummypox virus 3000 B.C.',
'gi|887699|gb|DQ37780 Squirrelpox virus 1296/99',
'gi|887699|gb|DQ37780 Squirrelpox virus 55',
], result)
def testTitleCollection(self):
"""
A title that occurs in the alignments of multiple reads must have
the data from both reads collected properly.
"""
mockOpener = mockOpen(read_data=(dumps(PARAMS) + '\n' +
dumps(RECORD2) + '\n' +
dumps(RECORD3) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
read2 = Read('id2', 'A' * 70)
read3 = Read('id3', 'A' * 70)
reads.add(read2)
reads.add(read3)
readsAlignments = BlastReadsAlignments(reads, 'file.json')
titlesAlignments = TitlesAlignments(readsAlignments)
title = 'gi|887699|gb|DQ37780 Cowpox virus 15'
titleAlignments = titlesAlignments[title]
self.assertEqual(title, titleAlignments.subjectTitle)
self.assertEqual(30000, titleAlignments.subjectLength)
self.assertEqual(2, len(titleAlignments))
self.assertEqual(read2, titleAlignments[0].read)
self.assertEqual(HSP(20), titleAlignments[0].hsps[0])
self.assertEqual(read3, titleAlignments[1].read)
self.assertEqual(HSP(20), titleAlignments[1].hsps[0])
def testExpectedTitleDetails(self):
"""
An instance of TitleAlignments in a TitlesAlignments instance must
have the expected attributes.
"""
mockOpener = mockOpen(read_data=(dumps(PARAMS) + '\n' +
dumps(RECORD0) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
read = Read('id0', 'A' * 70)
reads.add(read)
readsAlignments = BlastReadsAlignments(reads, 'file.json')
titlesAlignments = TitlesAlignments(readsAlignments)
title = 'gi|887699|gb|DQ37780 Squirrelpox virus 1296/99'
titleAlignments = titlesAlignments[title]
self.assertEqual(title, titleAlignments.subjectTitle)
self.assertEqual(37000, titleAlignments.subjectLength)
self.assertEqual(1, len(titleAlignments))
self.assertEqual(read, titleAlignments[0].read)
self.assertEqual(HSP(20), titleAlignments[0].hsps[0])
title = 'gi|887699|gb|DQ37780 Squirrelpox virus 55'
titleAlignments = titlesAlignments[title]
self.assertEqual(title, titleAlignments.subjectTitle)
self.assertEqual(38000, titleAlignments.subjectLength)
self.assertEqual(1, len(titleAlignments))
self.assertEqual(read, titleAlignments[0].read)
self.assertEqual(HSP(25), titleAlignments[0].hsps[0])
def testMaxScore_EValue(self):
"""
Sorting on max score must work when scores are e values, including a
secondary sort on title.
"""
mockOpener = mockOpen(
read_data=(dumps(PARAMS) + '\n' + dumps(RECORD0) + '\n' +
dumps(RECORD1) + '\n' + dumps(RECORD2) + '\n' +
dumps(RECORD3) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
reads.add(Read('id0', 'A' * 70))
reads.add(Read('id1', 'A' * 70))
reads.add(Read('id2', 'A' * 70))
reads.add(Read('id3', 'A' * 70))
readsAlignments = BlastReadsAlignments(
reads, 'file.json', scoreClass=LowerIsBetterScore)
titlesAlignments = TitlesAlignments(readsAlignments)
result = titlesAlignments.sortTitles('maxScore')
# self.assertEqual([
# 'gi|887699|gb|DQ37780 Cowpox virus 15', # 1e-6
# 'gi|887699|gb|DQ37780 Mummypox virus 3000 B.C.', # 1e-7
# 'gi|887699|gb|DQ37780 Monkeypox virus 456', # 1e-8
# 'gi|887699|gb|DQ37780 Squirrelpox virus 55', # 1e-10
# 'gi|887699|gb|DQ37780 Squirrelpox virus 1296/99', # 1e-11
# ], result)
self.assertEqual(
[
'gi|887699|gb|DQ37780 Squirrelpox virus 1296/99', # 1e-11
'gi|887699|gb|DQ37780 Squirrelpox virus 55', # 1e-10
'gi|887699|gb|DQ37780 Monkeypox virus 456', # 1e-8
'gi|887699|gb|DQ37780 Mummypox virus 3000 B.C.', # 1e-7
'gi|887699|gb|DQ37780 Cowpox virus 15', # 1e-6
],
result)
def testUnknown(self):
"""
Sorting on an unknown attribute must raise C{ValueError}.
"""
mockOpener = mockOpen(read_data=dumps(PARAMS) + '\n')
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
readsAlignments = BlastReadsAlignments(reads, 'file.json')
titlesAlignments = TitlesAlignments(readsAlignments)
self.assertRaises(ValueError, titlesAlignments.sortTitles, 'xxx')
def testEmpty(self):
"""
If passed an empty readsAlignments, titleCounts must return an
empty dictionary.
"""
mockOpener = mockOpen(read_data=dumps(PARAMS) + '\n')
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
readsAlignments = BlastReadsAlignments(reads, 'file.json')
self.assertEqual({}, titleCounts(readsAlignments))
def testEmpty(self):
"""
Sorting when there are no titles must return the empty list.
"""
mockOpener = mockOpen(read_data=dumps(PARAMS) + '\n')
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
readsAlignments = BlastReadsAlignments(reads, 'file.json')
titlesAlignments = TitlesAlignments(readsAlignments)
result = titlesAlignments.sortTitles('title')
self.assertEqual([], result)
def testEmpty(self):
"""
An instance of TitlesAlignments must have no titles if passed an
empty readsAlignments instance.
"""
mockOpener = mockOpen(read_data=(dumps(PARAMS) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
readsAlignments = BlastReadsAlignments(reads, 'file.json')
titlesAlignments = TitlesAlignments(readsAlignments)
self.assertEqual([], list(titlesAlignments.keys()))
def testDuplicatedTitle(self):
"""
If alignments for reads have a common title, the count on that title
must be correct.
"""
mockOpener = mockOpen(read_data=(dumps(PARAMS) + '\n' +
dumps(RECORD2) + '\n' +
dumps(RECORD3) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
reads.add(Read('id2', 'A' * 70))
reads.add(Read('id3', 'A' * 70))
readsAlignments = BlastReadsAlignments(reads, 'file.json')
self.assertEqual({
'gi|887699|gb|DQ37780 Cowpox virus 15': 2,
}, titleCounts(readsAlignments))
def testAddTitle(self):
"""
The addTitle function must add a title to the TitlesAlignments
instance.
"""
mockOpener = mockOpen(read_data=(dumps(PARAMS) + '\n' +
dumps(RECORD0) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
reads.add(Read('id0', 'A' * 70))
readsAlignments = BlastReadsAlignments(reads, 'file.json')
titlesAlignments = TitlesAlignments(readsAlignments)
title = 'gi|887699|gb|DQ37780 Squirrelpox virus 23'
titleAlignments = TitleAlignments(title, 55)
self.assertTrue(title not in titlesAlignments)
titlesAlignments.addTitle(title, titleAlignments)
self.assertTrue(title in titlesAlignments)
def testMaxTitlesZero(self):
"""
The filter function must return an empty result when maxTitles is zero.
"""
mockOpener = mockOpen(
read_data=(dumps(PARAMS) + '\n' + dumps(RECORD0) + '\n' +
dumps(RECORD1) + '\n' + dumps(RECORD2) + '\n' +
dumps(RECORD3) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
reads.add(Read('id0', 'A' * 70))
reads.add(Read('id1', 'A' * 70))
reads.add(Read('id2', 'A' * 70))
reads.add(Read('id3', 'A' * 70))
readsAlignments = BlastReadsAlignments(reads, 'file.json')
titlesAlignments = TitlesAlignments(readsAlignments)
result = titlesAlignments.filter(maxTitles=0, sortOn='maxScore')
self.assertEqual(0, len(result))
def testMaxTitlesNegative(self):
"""
The filter function must raise a ValueError if maxTitles is less than
zero.
"""
mockOpener = mockOpen(read_data=(dumps(PARAMS) + '\n' +
dumps(RECORD0) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
reads.add(Read('id0', 'A' * 70))
readsAlignments = BlastReadsAlignments(reads, 'file.json')
titlesAlignments = TitlesAlignments(readsAlignments)
error = '^maxTitles \(-1\) cannot be negative\.$'
six.assertRaisesRegex(self,
ValueError,
error,
titlesAlignments.filter,
maxTitles=-1)
def check(fastaFile, jsonFiles):
"""
Check for simple consistency between the FASTA file and the JSON files.
Note that some checking is already performed by the BlastReadsAlignments
class. That includes checking the number of reads matches the number of
BLAST records and that read ids and BLAST record read ids match.
@param jsonFiles: A C{list} of names of our BLAST JSON. These may
may be compressed (as bz2).
@param fastaFile: The C{str} name of a FASTA-containing file.
"""
reads = FastaReads(fastaFile)
readsAlignments = BlastReadsAlignments(reads, jsonFiles)
for index, readAlignments in enumerate(readsAlignments):
# Check that all the alignments in the BLAST JSON do not have query
# sequences or query offsets that are greater than the length of
# the sequence given in the FASTA file.
fastaLen = len(readAlignments.read)
for readAlignment in readAlignments:
for hsp in readAlignment.hsps:
# The FASTA sequence should be at least as long as the
# query in the JSON BLAST record (minus any gaps).
assert (fastaLen >=
len(hsp.query) - hsp.query.count('-')), (
'record %d: FASTA len %d < HSP query len %d.\n'
'FASTA: %s\nQuery match: %s' % (
index, fastaLen, len(hsp.query),
readAlignments.read.sequence, hsp.query))
# The FASTA sequence length should be larger than either of
# the query offsets mentioned in the JSON BLAST
# record. That's because readStart and readEnd are offsets
# into the read - so they can't be bigger than the read
# length.
#
# TODO: These asserts should be more informative when they
# fail.
assert fastaLen >= hsp.readEnd >= hsp.readStart, (
'record %d: FASTA len %d not greater than both read '
'offsets (%d - %d), or read offsets are non-increasing. '
'FASTA: %s\nQuery match: %s' % (
index, fastaLen, hsp.readStart, hsp.readEnd,
readAlignments.read.sequence, hsp.query))
def testAddTitleRepeat(self):
"""
The addTitle function must raise a C{KeyError} if an attempt is made
to add a pre-existing title to a TitlesAlignments instance.
"""
mockOpener = mockOpen(read_data=(dumps(PARAMS) + '\n' +
dumps(RECORD0) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
reads.add(Read('id0', 'A' * 70))
readsAlignments = BlastReadsAlignments(reads, 'file.json')
titlesAlignments = TitlesAlignments(readsAlignments)
title = 'gi|887699|gb|DQ37780 Squirrelpox virus 1296/99'
titleAlignments = TitleAlignments(title, 55)
error = ("Title 'gi\|887699\|gb\|DQ37780 Squirrelpox virus "
"1296/99' already present in TitlesAlignments instance\.")
six.assertRaisesRegex(self, KeyError, error,
titlesAlignments.addTitle, title,
titleAlignments)
def testCoverageExcludesAll(self):
"""
The coverage function must return an titlesAlignments instance with
no titles if none of its titles has sufficient coverage.
"""
mockOpener = mockOpen(
read_data=(dumps(PARAMS) + '\n' + dumps(RECORD0) + '\n' +
dumps(RECORD1) + '\n' + dumps(RECORD2) + '\n' +
dumps(RECORD3) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
reads.add(Read('id0', 'A' * 70))
reads.add(Read('id1', 'A' * 70))
reads.add(Read('id2', 'A' * 70))
reads.add(Read('id3', 'A' * 70))
readsAlignments = BlastReadsAlignments(reads, 'file.json')
titlesAlignments = TitlesAlignments(readsAlignments)
result = titlesAlignments.filter(minCoverage=0.1)
self.assertEqual(0, len(result))
def testHsps(self):
"""
The hsps function must yield all the hsps for all titles in a
TitlesAlignments instance.
"""
mockOpener = mockOpen(
read_data=(dumps(PARAMS) + '\n' + dumps(RECORD0) + '\n' +
dumps(RECORD1) + '\n' + dumps(RECORD2) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
reads.add(Read('id0', 'A' * 70))
reads.add(Read('id1', 'A' * 70))
reads.add(Read('id2', 'A' * 70))
readsAlignments = BlastReadsAlignments(reads, 'file.json')
titlesAlignments = TitlesAlignments(readsAlignments)
result = list(titlesAlignments.hsps())
self.assertEqual(
sorted([HSP(20), HSP(25),
HSP(20), HSP(20),
HSP(20)]), sorted(result))
def testUnknownSortOn(self):
"""
The filter function must raise a ValueError if the passed sortOn
value isn't recognized.
"""
mockOpener = mockOpen(read_data=(dumps(PARAMS) + '\n' +
dumps(RECORD0) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
reads.add(Read('id0', 'A' * 70))
readsAlignments = BlastReadsAlignments(reads, 'file.json')
titlesAlignments = TitlesAlignments(readsAlignments)
error = ('^Sort attribute must be one of "length", "maxScore", '
'"medianScore", "readCount", "title"\.$')
six.assertRaisesRegex(self,
ValueError,
error,
titlesAlignments.filter,
maxTitles=0,
sortOn='unknown')
def testExpectedTitles(self):
"""
An instance of TitlesAlignments must have the expected titles.
"""
mockOpener = mockOpen(
read_data=(dumps(PARAMS) + '\n' + dumps(RECORD0) + '\n' +
dumps(RECORD1) + '\n' + dumps(RECORD2) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
reads.add(Read('id0', 'A' * 70))
reads.add(Read('id1', 'A' * 70))
reads.add(Read('id2', 'A' * 70))
readsAlignments = BlastReadsAlignments(reads, 'file.json')
titlesAlignments = TitlesAlignments(readsAlignments)
self.assertEqual([
'gi|887699|gb|DQ37780 Cowpox virus 15',
'gi|887699|gb|DQ37780 Monkeypox virus 456',
'gi|887699|gb|DQ37780 Mummypox virus 3000 B.C.',
'gi|887699|gb|DQ37780 Squirrelpox virus 1296/99',
'gi|887699|gb|DQ37780 Squirrelpox virus 55',
], sorted(titlesAlignments.keys()))
def testMinMatchingReads(self):
"""
The filter function work correctly when passed a value for
minMatchingReads.
"""
mockOpener = mockOpen(
read_data=(dumps(PARAMS) + '\n' + dumps(RECORD0) + '\n' +
dumps(RECORD1) + '\n' + dumps(RECORD2) + '\n' +
dumps(RECORD3) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
reads.add(Read('id0', 'A' * 70))
reads.add(Read('id1', 'A' * 70))
reads.add(Read('id2', 'A' * 70))
reads.add(Read('id3', 'A' * 70))
readsAlignments = BlastReadsAlignments(reads, 'file.json')
titlesAlignments = TitlesAlignments(readsAlignments)
result = titlesAlignments.filter(minMatchingReads=2)
self.assertEqual([
'gi|887699|gb|DQ37780 Cowpox virus 15',
], list(result.keys()))
def testWithScoreBetterThan_Bits(self):
"""
The filter function work correctly when passed a value for
withScoreBetterThan when using bit scores.
"""
mockOpener = mockOpen(
read_data=(dumps(PARAMS) + '\n' + dumps(RECORD0) + '\n' +
dumps(RECORD1) + '\n' + dumps(RECORD2) + '\n' +
dumps(RECORD3) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
reads.add(Read('id0', 'A' * 70))
reads.add(Read('id1', 'A' * 70))
reads.add(Read('id2', 'A' * 70))
reads.add(Read('id3', 'A' * 70))
readsAlignments = BlastReadsAlignments(reads, 'file.json')
titlesAlignments = TitlesAlignments(readsAlignments)
result = titlesAlignments.filter(withScoreBetterThan=24)
self.assertEqual([
'gi|887699|gb|DQ37780 Squirrelpox virus 55',
], list(result.keys()))
def testMaxTitlesOne(self):
"""
The filter function must return just the best title when maxTitles
is one.
"""
mockOpener = mockOpen(
read_data=(dumps(PARAMS) + '\n' + dumps(RECORD0) + '\n' +
dumps(RECORD1) + '\n' + dumps(RECORD2) + '\n' +
dumps(RECORD3) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
reads.add(Read('id0', 'A' * 70))
reads.add(Read('id1', 'A' * 70))
reads.add(Read('id2', 'A' * 70))
reads.add(Read('id3', 'A' * 70))
readsAlignments = BlastReadsAlignments(reads, 'file.json')
titlesAlignments = TitlesAlignments(readsAlignments)
result = titlesAlignments.filter(maxTitles=1, sortOn='maxScore')
self.assertEqual([
'gi|887699|gb|DQ37780 Squirrelpox virus 55',
], sorted(result.keys()))
def testTwoJSONInputsWithSubjectInCommon(self):
"""
If two JSON files are passed to L{BlastReadsAlignments} with a matched
subject in common and a TitlesAlignments is made, the title in the
TitlesAlignments must have information from both reads, including the
correct HSP scores.
"""
class SideEffect(object):
def __init__(self):
self.first = True
def sideEffect(self, _ignoredFilename, **kwargs):
if self.first:
self.first = False
return File([dumps(PARAMS) + '\n', dumps(RECORD2) + '\n'])
else:
return File([dumps(PARAMS) + '\n', dumps(RECORD4) + '\n'])
title = 'gi|887699|gb|DQ37780 Cowpox virus 15'
sideEffect = SideEffect()
with patch.object(builtins, 'open') as mockMethod:
mockMethod.side_effect = sideEffect.sideEffect
reads = Reads()
reads.add(Read('id2', 'A' * 70))
reads.add(Read('id4', 'A' * 70))
readsAlignments = BlastReadsAlignments(
reads, ['file1.json', 'file2.json'])
titlesAlignments = TitlesAlignments(readsAlignments)
titleAlignments = titlesAlignments[title]
self.assertEqual(title, titleAlignments.subjectTitle)
self.assertEqual(4, titleAlignments.hspCount())
self.assertEqual('id2', titleAlignments[0].read.id)
self.assertEqual('id4', titleAlignments[1].read.id)
# First matching read has one HSP.
self.assertEqual(HSP(20), titleAlignments[0].hsps[0])
# Second matching read has three HSPs.
self.assertEqual(HSP(10), titleAlignments[1].hsps[0])
self.assertEqual(HSP(5), titleAlignments[1].hsps[1])
self.assertEqual(HSP(3), titleAlignments[1].hsps[2])
def testThreeRecords(self):
"""
If alignments for three reads are passed to titleCounts, it must
return the correct title counts.
"""
mockOpener = mockOpen(
read_data=(dumps(PARAMS) + '\n' + dumps(RECORD0) + '\n' +
dumps(RECORD1) + '\n' + dumps(RECORD2) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
reads.add(Read('id0', 'A' * 70))
reads.add(Read('id1', 'A' * 70))
reads.add(Read('id2', 'A' * 70))
readsAlignments = BlastReadsAlignments(reads, 'file.json')
self.assertEqual(
{
'gi|887699|gb|DQ37780 Squirrelpox virus 1296/99': 1,
'gi|887699|gb|DQ37780 Mummypox virus 3000 B.C.': 1,
'gi|887699|gb|DQ37780 Cowpox virus 15': 1,
'gi|887699|gb|DQ37780 Monkeypox virus 456': 1,
'gi|887699|gb|DQ37780 Squirrelpox virus 55': 1
}, titleCounts(readsAlignments))
def testMaxTitlesTwoSortOnLength(self):
"""
The filter function must return the two titles whose sequences are the
longest when maxTitles is 2 and sortOn is 'length'.
"""
mockOpener = mockOpen(
read_data=(dumps(PARAMS) + '\n' + dumps(RECORD0) + '\n' +
dumps(RECORD1) + '\n' + dumps(RECORD2) + '\n' +
dumps(RECORD3) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
reads.add(Read('id0', 'A' * 70))
reads.add(Read('id1', 'A' * 70))
reads.add(Read('id2', 'A' * 70))
reads.add(Read('id3', 'A' * 70))
readsAlignments = BlastReadsAlignments(reads, 'file.json')
titlesAlignments = TitlesAlignments(readsAlignments)
result = titlesAlignments.filter(maxTitles=2, sortOn='length')
self.assertEqual([
'gi|887699|gb|DQ37780 Squirrelpox virus 1296/99',
'gi|887699|gb|DQ37780 Squirrelpox virus 55',
], sorted(result.keys()))
def testReadSetFilterStrict(self):
"""
The filter function work correctly when passed a 1.0 value for
minNewReads.
"""
mockOpener = mockOpen(
read_data=(dumps(PARAMS) + '\n' + dumps(RECORD0) + '\n' +
dumps(RECORD1) + '\n' + dumps(RECORD2) + '\n' +
dumps(RECORD3) + '\n'))
with patch.object(builtins, 'open', mockOpener):
reads = Reads()
reads.add(Read('id0', 'A' * 70))
reads.add(Read('id1', 'A' * 70))
reads.add(Read('id2', 'A' * 70))
reads.add(Read('id3', 'A' * 70))
readsAlignments = BlastReadsAlignments(reads, 'file.json')
titlesAlignments = TitlesAlignments(readsAlignments)
result = titlesAlignments.filter(minNewReads=1.0)
# Either 'gi|887699|gb|DQ37780 Mummypox virus 3000 B.C.'
# invalidates 'gi|887699|gb|DQ37780 Monkeypox virus 456' or
# vice-versa. It depends on Python's dict walking order. Check
# for both, making sure just one of them is true.
mummypox = 'gi|887699|gb|DQ37780 Mummypox virus 3000 B.C.'
monkeypox = 'gi|887699|gb|DQ37780 Monkeypox virus 456'
assertionCount = 0
if mummypox in result:
self.assertTrue(
monkeypox in result.readSetFilter.invalidates(mummypox))
assertionCount += 1
if monkeypox in result:
self.assertTrue(
mummypox in result.readSetFilter.invalidates(monkeypox))
assertionCount += 1
self.assertEqual(1, assertionCount)
jsonFiles = list(chain.from_iterable(args.json))
whitelist = (
set(chain.from_iterable(args.whitelist)) if args.whitelist else None)
blacklist = (
set(chain.from_iterable(args.blacklist)) if args.blacklist else None)
# TODO: Add a --readClass command-line option in case we want to
# process FASTA containing AA sequences.
if args.fasta:
reads = FastaReads(list(chain.from_iterable(args.fasta)))
else:
reads = FastqReads(list(chain.from_iterable(args.fastq)))
if args.matcher == 'blast':
from dark.blast.alignments import BlastReadsAlignments
readsAlignments = BlastReadsAlignments(reads, jsonFiles)
else:
# Must be 'diamond' (due to parser.add_argument 'choices' argument).
if (args.diamondDatabaseFastaFilename is None and
args.diamondSqliteDatabaseFilename is None):
print('Either --diamondDatabaseFastaFilename or '
'--diamondSqliteDatabaseFilename must be used with '
'--matcher diamond.', file=sys.stderr)
sys.exit(1)
elif not (args.diamondDatabaseFastaFilename is None or
args.diamondSqliteDatabaseFilename is None):
print('--diamondDatabaseFastaFilename and '
'--diamondSqliteDatabaseFilename cannot both be used with '
'--matcher diamond.', file=sys.stderr)
sys.exit(1)
