def testCountGenomesEmpty(self):
"""
An empty database must have zero genomes.
"""
writer = SqliteIndexWriter(':memory:')
db = SqliteIndex(writer._connection)
self.assertEqual(0, db.genomeCount())
writer.close()
db.close()
def testCountProteinsOne(self):
"""
A database with one protein must have a protein count of one.
"""
writer = SqliteIndexWriter(':memory:')
writer.addProtein('NC222', 'NC23', 'AAA', 'offets', True, False, 3)
db = SqliteIndex(writer._connection)
self.assertEqual(1, db.proteinCount())
writer.close()
db.close()
def main(args, parser):
"""
Build the protein database.
@param args: The namespace of command-line arguments returned by
argparse.parse_args()
@param parser: An C{argparse.ArgumentParser} instance.
"""
if (args.minGenomeLength is not None and args.maxGenomeLength is not None
and args.minGenomeLength > args.maxGenomeLength):
raise ValueError(
'--minGenomeLength cannot be larger than --maxGenomeLength')
if args.excludeExclusiveHost:
excludeExclusiveHosts = set(
chain.from_iterable(args.excludeExclusiveHost))
else:
excludeExclusiveHosts = None
taxonomyDatabase = parseTaxonomyDatabaseCommandLineOptions(args, parser)
progress = args.progress
if progress:
overallStart = time()
totalGenomeCount = totalProteinCount = 0
with SqliteIndexWriter(args.databaseFile) as db:
for fileCount, (filename, addFunc,
type_) in enumerate(filenamesAndAdders(args, db),
start=1):
if args.logFile:
print("\n>>> Indexing '%s'." % filename,
end='\n\n',
file=args.logFile)
if progress:
start = time()
examinedGenomeCount, genomeCount, proteinCount = addFunc(
filename,
dnaOnly=args.dnaOnly,
rnaOnly=args.rnaOnly,
minGenomeLength=args.minGenomeLength,
maxGenomeLength=args.maxGenomeLength,
excludeExclusiveHosts=excludeExclusiveHosts,
excludeFungusOnlyViruses=args.excludeFungusOnlyViruses,
excludePlantOnlyViruses=args.excludePlantOnlyViruses,
databaseName=args.databaseName,
taxonomyDatabase=taxonomyDatabase,
proteinSource=args.proteinSource,
genomeSource=args.genomeSource,
duplicationPolicy=args.duplicationPolicy,
logfp=args.logFile)
if examinedGenomeCount == 0:
if type_ == 'gb':
print('WARNING: No genomes found in %r. Did the GenBank '
'download fail on that file?' % filename,
file=sys.stderr)
else:
assert type_ == 'json'
print('WARNING: no genomes found in JSON file %r.' %
filename,
file=sys.stderr)
if progress:
elapsed = time() - start
totalGenomeCount += genomeCount
totalProteinCount += proteinCount
print('Processed %r: added %3d of %3d genome%s (%5d '
'protein%s) in %.2f seconds.' %
(filename, genomeCount, examinedGenomeCount,
' ' if examinedGenomeCount == 1 else 's', proteinCount,
'' if proteinCount == 1 else 's', elapsed),
file=sys.stderr)
if progress:
elapsed = time() - overallStart
print('%d files (containing %d genomes and %d proteins) '
'indexed in %.2f seconds (%.2f mins).' %
(fileCount, totalGenomeCount, totalProteinCount, elapsed,
elapsed / 60),
file=sys.stderr)
def testTibetanFrogHBV(self):
"""
Test that Tibetan frogs can get HBV.
"""
proteinAccession = 'YP_009259545.1'
proteinSequence = SAMPLE_DATA['proteins'][proteinAccession]['protein']
proteinId = SAMPLE_DATA['proteins'][proteinAccession]['id']
proteinRange = SAMPLE_DATA['proteins'][proteinAccession]['range']
ranges = GenomeRanges(proteinRange)
queryStartInProtein = 10 # This is a 0-based amino acid offset.
queryLenInProtein = 40 # This is an amino acid length.
genomeAccession = 'NC_030446.1'
genomeSequence = SAMPLE_DATA['genomes'][genomeAccession]['genome']
genomeLen = len(genomeSequence)
# The query sequence is nucleotides that match the amino acids in the
# protein. Here we use the first ([0] in the below) codon for each
# amino acid to make the nucleotide sequence.
queryId = 'query'
querySequence = ''.join(
CODONS[aa][0]
for aa in proteinSequence[queryStartInProtein:queryStartInProtein +
queryLenInProtein])
queryQuality = 'E' * len(querySequence)
# Use the protein sequence to make a DIAMOND database and run DIAMOND
# on the query. Yes, this really runs DIAMOND, so you need to have it
# installed, with its executable somewhere in your shell's PATH.
with DiamondExecutor() as de:
de.addSubject(Read(proteinId, proteinSequence))
queries = Reads([Read(queryId, querySequence, queryQuality)])
(diamondResult, ) = list(de.search(queries))
# Make sure DIAMOND gives us back what we expected.
self.assertEqual(
{
'bitscore': 83.6,
'btop': str(queryLenInProtein), # Exact match of all AAs.
'qframe': 1,
'qend': 3 * queryLenInProtein,
'full_qqual': queryQuality,
'qlen': len(querySequence),
'full_qseq': querySequence,
'qseqid': 'query',
'qstart': 1,
'slen': len(proteinSequence),
'sstart': queryStartInProtein + 1, # DIAMOND is 1-based.
'stitle': proteinId,
},
diamondResult)
# Make a genomes/proteins sqlite database and add information about
# the protein and the nucleotide genome it comes from.
db = SqliteIndexWriter(':memory:')
db.addProtein(proteinAccession, genomeAccession, proteinSequence,
proteinRange, True, ranges.circular(genomeLen),
ranges.distinctRangeCount(genomeLen))
genome = _Genome(
# genomeAccession, genomeName, genomeSequence
)
db.addGenome(genome, 1, 'test-db')
# Make a DIAMOND-to-SAM writer and give it the DIAMOND output.
writer = SimpleDiamondSAMWriter(SqliteIndex(db._connection))
writer.addMatch('\t'.join(
map(str, (diamondResult['bitscore'], diamondResult['btop'],
diamondResult['qframe'], diamondResult['qend'],
diamondResult['full_qqual'], diamondResult['qlen'],
diamondResult['full_qseq'], diamondResult['qseqid'],
diamondResult['qstart'], diamondResult['slen'],
diamondResult['sstart'], diamondResult['stitle']))))
# Tell the writer to save the matches as SAM and check the result.
fp = StringIO()
writer.save(filename=fp)
flags = '31'
self.assertEqual(
'\n'.join([
'@SQ\tSN:%s\tLN:%d' % (genomeAccession, len(genomeSequence)),
'\t'.join([
queryId,
'0',
genomeAccession,
flags,
'255',
'120M', # (Exact) match of 40 AAs.
'*',
'0',
'0',
querySequence,
queryQuality,
'AS:i:%d' % int(diamondResult['bitscore']),
])
]) + '\n',
fp.getvalue())
def testPolymerase(self):
"""
Test the polymerase protein.
"""
self.maxDiff = None
proteinAccession = 'CAK55121.1'
proteinSequence = SAMPLE_DATA['proteins'][proteinAccession]['protein']
proteinId = SAMPLE_DATA['proteins'][proteinAccession]['id']
proteinRange = SAMPLE_DATA['proteins'][proteinAccession]['range']
ranges = GenomeRanges(proteinRange)
queryStartInProtein = 270 # This is a 0-based amino acid offset.
queryLenInProtein = 470 # This is an amino acid length.
genomeAccession = 'AM282986.1'
genomeId = SAMPLE_DATA['genomes'][genomeAccession]['id']
genomeSequence = SAMPLE_DATA['genomes'][genomeAccession]['genome']
genomeLen = len(genomeSequence)
# The query sequence is nucleotides that match the amino acids in the
# protein. Here we use the first ([0] in the below) codon for each
# amino acid to make the nucleotide sequence.
queryId = 'query'
querySequence = ''.join(
CODONS[aa][0]
for aa in proteinSequence[queryStartInProtein:queryStartInProtein +
queryLenInProtein])
queryQuality = 'E' * len(querySequence)
# Use the protein sequence to make a DIAMOND database and run DIAMOND
# on the query. Yes, this really runs DIAMOND, so you need to have it
# installed, with its executable somewhere in your shell's PATH.
with DiamondExecutor() as de:
de.addSubject(Read(proteinId, proteinSequence))
queries = Reads([Read(queryId, querySequence, queryQuality)])
(diamondResult, ) = list(de.search(queries))
# Make sure DIAMOND gives us back what we expected.
self.assertEqual(
{
'bitscore': 974.0,
'btop': str(queryLenInProtein), # Exact match of all query AA.
'qframe': 1,
'qstart': 1,
'qend': 3 * queryLenInProtein,
'qlen': len(querySequence),
'qseqid': 'query',
'full_qseq': querySequence,
'full_qqual': queryQuality,
'slen': len(proteinSequence),
'sstart': queryStartInProtein + 1, # DIAMOND is 1-based.
'stitle': proteinId,
},
diamondResult)
# Make a genomes/proteins sqlite database and add information about
# the protein and the nucleotide genome it comes from.
db = SqliteIndexWriter(':memory:')
db.addProtein(proteinAccession, genomeAccession, proteinSequence,
proteinRange, True, ranges.circular(genomeLen),
ranges.distinctRangeCount(genomeLen))
genome = _Genome({
'id': genomeAccession,
'name': genomeId,
'sequence': genomeSequence,
'features': [],
})
db.addGenome(genome,
source={
'host': 'H**o sapiens',
'mol_type': 'DNA',
'organism': 'Hepatitis B Virus',
},
taxonomyId=500,
proteinCount=len(SAMPLE_DATA['proteins']),
databaseName='test-db')
# Make a DIAMOND-to-SAM writer and give it the DIAMOND output.
writer = SimpleDiamondSAMWriter(SqliteIndex(db._connection))
writer.addMatch('\t'.join(
map(str, (diamondResult['bitscore'], diamondResult['btop'],
diamondResult['qframe'], diamondResult['qend'],
diamondResult['full_qqual'], diamondResult['qlen'],
diamondResult['full_qseq'], diamondResult['qseqid'],
diamondResult['qstart'], diamondResult['slen'],
diamondResult['sstart'], diamondResult['stitle']))))
# Tell the writer to save the matches as SAM and check the result.
fp = StringIO()
writer.save(filename=fp)
flags = 0
self.assertEqual(
'\n'.join(('@SQ\tSN:%s\tLN:%d' %
(genomeAccession, len(genomeSequence)), '\t'.join(
map(str, (
queryId,
flags,
genomeAccession,
queryStartInProtein * 3 + 1,
255,
'1410M',
'*',
0,
0,
querySequence,
queryQuality,
'AS:i:%d' % int(diamondResult['bitscore']),
))))) + '\n', fp.getvalue())