def test_bulk_index_kmers(dna_kmer_index):
A = dna_kmer_index.db.alphabet
dna_kmer_index.db.initialize()
S = A.parse('ATGCA', name='foo')
T = A.parse('ATGCC', name='bar')
fasta = StringIO()
write_fasta(fasta, [S, T])
fasta.seek(0)
S_rec, T_rec = dna_kmer_index.db.load_fasta(fasta)
assert S_rec is not None and T_rec is not None
assert dna_kmer_index.num_kmers() == 4
assert dna_kmer_index.total_length_indexed() == len(S) + len(T)
# find the occurences of 'ATG'
S_id, T_id = S_rec.id, T_rec.id
atg_int = dna_kmer_index.kmer_as_int((0, 3, 2))
atg_hits = [(hits, score) for kmer, hits, score in dna_kmer_index.kmers()
if kmer == atg_int]
assert len(atg_hits) == 1 and atg_hits[0][0] == [(S_id, 0), (T_id, 0)]
dna_kmer_index.score_kmers()
atg_hits = [(hits, score) for kmer, hits, score in dna_kmer_index.kmers()
if kmer == atg_int]
atg_score = atg_hits[0][1]
assert atg_score is not None and score > 0
# ATG is the most common kmer, it most have the highest score:
assert all(atg_score >= score for _, _, score in dna_kmer_index.kmers())
# shouldn't do anything
dna_kmer_index.score_kmers(only_missing=True)
def test_database_populate_fasta():
A = Alphabet('ACGT')
S = A.parse('AACT', name='S')
T = A.parse('GCAT', name='T')
db = DB(':memory:', A)
db.initialize()
fasta = StringIO()
fasta.name = '/x.fasta'
write_fasta(fasta, [S, T])
fasta.seek(0)
inserted = db.load_fasta(fasta, rc=False)
assert len(inserted) == 2
assert all(isinstance(r, Record) for r in inserted)
assert all(rec.source_file == fasta.name for rec in inserted), \
'source file of sequence records must be set'
assert [db.load_from_record(rec, fasta) for rec in inserted] == [S, T], \
'should be able to retrieve sequences by position in source'
with patch('biseqt.database.open', create=True) as open_mock:
open_mock.return_value = MagicMock(spec=file, wraps=fasta)
assert db.load_from_record(inserted[0]) == S, \
'load_from_record should work without an open file handle'
def test_write_fasta():
A = Alphabet('ACGT')
S = A.parse('AAA', name='foo')
T = A.parse('TTT', name='bar')
with NamedTemporaryFile() as f:
write_fasta(f, [S, T])
f.seek(0)
assert [s for s, _ in read_fasta(f, A)] == [S, T], \
'read_fasta(write_fasta()) should be identity'
f = StringIO('')
write_fasta(f, [S, T])
f.seek(0)
assert f.read() == '>foo\nAAA\n>bar\nTTT\n', 'should work on StringIO'
f = StringIO('')
# duplicate names not allowed
with pytest.raises(AssertionError):
write_fasta(f, [S, S])
f = StringIO('')
S = A.parse('AAATTT', name='foo')
write_fasta(f, [S], width=3) # should take 3 lines
f.seek(0)
assert sum(1 for _ in f) == 3, 'FASTA width should be modifiable'
def seed_index():
"""Creates a database, a kmer index, and a seed index with word length 5
stored in memory and returns the seed index. The database is populated with
3 random sequences of length 100 and all kmers and seeds are indexed."""
A = Alphabet('ACGT')
num_seqs = 3
seq_len = 100
wordlen = 5
db = DB(':memory:', A)
seed_index = SeedIndex(KmerIndex(db, wordlen))
seed_index.db.initialize()
fasta = StringIO()
seqs = (rand_seq(A, seq_len).to_named('#%d' % i) for i in range(num_seqs))
write_fasta(fasta, seqs)
fasta.seek(0)
db.load_fasta(fasta)
seed_index.index_seeds()
return seed_index
def test_database_populate_fasta_rc():
A = Alphabet('ACGT')
S = A.parse('AACT', name='S')
T = A.parse('GCAT', name='T')
db = DB(':memory:', A)
db.initialize()
fasta = StringIO()
write_fasta(fasta, [S, T])
fasta.seek(0)
inserted = db.load_fasta(fasta, rc=True)
assert len(inserted) == 4
assert [r.attrs['rc_of'] for r in inserted if 'rc_of' in r.attrs] \
== [S.content_id, T.content_id], \
'reverse complements should know what their origin is'
def cond_T_rc(r): return r.attrs.get('rc_of', None) == T.content_id
found_T_rc = next(db.find(condition=cond_T_rc))
T_rc = T.reverse().transform(['AT', 'CG'], name='(rc) ' + T.name)
assert db.load_from_record(found_T_rc, fasta) == T_rc, \
'reverse complements should load properly from a record'