def test_pseudo_count(self):
# use example from
# http://biologie.univ-mrs.fr/upload/p202/01.4.PSSM_theory.pdf
alpha = unambiguous_dna
dna_align = MultipleSeqAlignment([
SeqRecord(Seq("AACCACGTTTAA", alpha), id="ID001"),
SeqRecord(Seq("CACCACGTGGGT", alpha), id="ID002"),
SeqRecord(Seq("CACCACGTTCGC", alpha), id="ID003"),
SeqRecord(Seq("GCGCACGTGGGG", alpha), id="ID004"),
SeqRecord(Seq("TCGCACGTTGTG", alpha), id="ID005"),
SeqRecord(Seq("TGGCACGTGTTT", alpha), id="ID006"),
SeqRecord(Seq("TGACACGTGGGA", alpha), id="ID007"),
SeqRecord(Seq("TTACACGTGCGC", alpha), id="ID008")
])
summary = SummaryInfo(dna_align)
expected = FreqTable({
"A": 0.325,
"G": 0.175,
"T": 0.325,
"C": 0.175
}, FREQ, unambiguous_dna)
ic = summary.information_content(e_freq_table=expected,
log_base=math.exp(1),
pseudo_count=1)
self.assertAlmostEqualList(summary.ic_vector, [
0.110, 0.090, 0.360, 1.290, 0.800, 1.290, 1.290, 0.80, 0.610,
0.390, 0.470, 0.040
],
places=2)
self.assertAlmostEqual(ic, 7.546, places=3)
def validate_msa_characters(msa, legal_chars="-acgtn"):
msa_info = SummaryInfo(msa)
msa_letters = msa_info._get_all_letters()
for let in msa_letters:
if let not in legal_chars:
return False
return True
def test_proteins(self):
alpha = HasStopCodon(Gapped(generic_protein, "-"), "*")
a = MultipleSeqAlignment([
SeqRecord(Seq("MHQAIFIYQIGYP*LKSGYIQSIRSPEYDNW-", alpha),
id="ID001"),
SeqRecord(Seq("MH--IFIYQIGYAYLKSGYIQSIRSPEY-NW*", alpha),
id="ID002"),
SeqRecord(Seq("MHQAIFIYQIGYPYLKSGYIQSIRSPEYDNW*", alpha),
id="ID003")
])
self.assertEqual(32, a.get_alignment_length())
s = SummaryInfo(a)
c = s.dumb_consensus(ambiguous="X")
self.assertEqual(str(c), "MHQAIFIYQIGYXXLKSGYIQSIRSPEYDNW*")
c = s.gap_consensus(ambiguous="X")
self.assertEqual(str(c), "MHXXIFIYQIGYXXLKSGYIQSIRSPEYXNWX")
m = s.pos_specific_score_matrix(chars_to_ignore=['-', '*'], axis_seq=c)
self.assertEqual(
str(m),
""" A D E F G H I K L M N P Q R S W Y
M 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
H 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
X 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.0 0.0 0.0 0.0 0.0
X 2.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
I 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
F 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
I 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Y 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0
Q 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0
I 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
G 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Y 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0
X 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.0 0.0 0.0 0.0 0.0 0.0
X 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.0
L 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
K 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
S 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0
G 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Y 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0
I 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Q 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0
S 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0
I 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
R 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0
S 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0
P 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0
E 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Y 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0
X 0.0 2.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
N 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0
W 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0
X 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
""")
ic = s.information_content(chars_to_ignore=['-', '*'])
self.assertAlmostEqual(ic, 133.061475107, places=6)
def calcAbundance(alignmentFile, consensusFile, abundanceFile, abundancePercentFile, verbose):
print('Calculating the abundance matrix...')
alignment = AlignIO.read(alignmentFile, "fasta")
summary = SummaryInfo(alignment)
consensusSeq = SeqIO.read(consensusFile, 'fasta')
if (len(consensusSeq) == alignment.get_alignment_length()):
abundanceMatrix = summary.pos_specific_score_matrix(consensusSeq)
else:
with open(consensusFile, "w") as f:
SeqIO.write(SeqRecord( summary.dumb_consensus(), id='consensus'), f, "fasta")
abundanceMatrix = summary.pos_specific_score_matrix()
if verbose:
print("Abundance matrix (absolute values):")
print(str(abundanceMatrix))
with open(abundanceFile, 'w') as f:
f.write(str(abundanceMatrix))
for pos, abundance in enumerate(abundanceMatrix):
for res, value in abundance.items():
abundanceMatrix[pos][res] = 100.0 * float(value) / float(len(alignment))
if verbose:
print("Abundance matrix (percentages):")
print(str(abundanceMatrix))
with open(abundancePercentFile, 'w') as f:
f.write(str(abundanceMatrix))
print('OK')
def test_proteins(self):
a = MultipleSeqAlignment([
SeqRecord(Seq("MHQAIFIYQIGYP*LKSGYIQSIRSPEYDNW-"), id="ID001"),
SeqRecord(Seq("MH--IFIYQIGYAYLKSGYIQSIRSPEY-NW*"), id="ID002"),
SeqRecord(Seq("MHQAIFIYQIGYPYLKSGYIQSIRSPEYDNW*"), id="ID003")
])
self.assertEqual(32, a.get_alignment_length())
s = SummaryInfo(a)
c = s.dumb_consensus(ambiguous="X")
self.assertEqual(str(c), "MHQAIFIYQIGYXXLKSGYIQSIRSPEYDNW*")
c = s.gap_consensus(ambiguous="X")
self.assertEqual(str(c), "MHXXIFIYQIGYXXLKSGYIQSIRSPEYXNWX")
m = s.pos_specific_score_matrix(chars_to_ignore=["-", "*"], axis_seq=c)
self.assertEqual(
str(m),
""" A D E F G H I K L M N P Q R S W Y
M 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
H 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
X 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.0 0.0 0.0 0.0 0.0
X 2.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
I 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
F 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
I 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Y 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0
Q 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0
I 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
G 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Y 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0
X 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.0 0.0 0.0 0.0 0.0 0.0
X 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.0
L 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
K 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
S 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0
G 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Y 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0
I 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Q 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0
S 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0
I 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
R 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0
S 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0
P 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0
E 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Y 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0
X 0.0 2.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
N 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0
W 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 0.0
X 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
""")
letters = IUPACData.protein_letters
base_freq = 1.0 / len(letters)
e_freq_table = {letter: base_freq for letter in letters}
ic = s.information_content(e_freq_table=e_freq_table,
chars_to_ignore=["-", "*"])
self.assertAlmostEqual(ic, 133.061475107, places=6)
def add_consensus(alignment, threshold=0.9, ambiguous='-', name='consensus'):
"""Add a consensus line"""
a = SummaryInfo(alignment)
# cons=a.dumb_consensus(threshold, ambiguous)
cons = a.gap_consensus(threshold, ambiguous)
alignment.extend([SeqRecord(cons, id=name, name=name)])
return alignment
def trim_aln_to_seq(alignment, sequence):
"""Trim alignment to a sequence, i.e. leave only postions that correspond to this sequence
Note that seqeuence should be incorportatable into alignment without additional gaps in alignment.
"""
n1 = str(uuid.uuid4())
n2 = str(uuid.uuid4())
#Get consensus
a = SummaryInfo(alignment)
cons = a.dumb_consensus(threshold=0.1, ambiguous='X')
#Needle it
SeqIO.write([SeqRecord(cons, id='CONS', name='CONS')], n1 + '.fasta',
'fasta')
SeqIO.write([SeqRecord(sequence, id='KEY', name='KEY')], n2 + '.fasta',
'fasta')
#Now we will redo it with Needlman Wunsh - the global alignment
needle_cline = NeedleCommandline(asequence=n1 + ".fasta",
bsequence=n2 + ".fasta",
gapopen=10,
gapextend=0.5,
outfile=n1 + ".txt")
stdout, stderr = needle_cline()
# print('Needle alignment')
align = AlignIO.read(n1 + ".txt", "emboss")
os.system('rm %s.fasta %s.fasta %s.txt' % (n1, n2, n1))
# print align
# print alignment
align.extend(alignment)
a = align[1:, :]
return trim_aln_to_key_seq(a, sequence)[1:, :]
def trim_aln_to_seq_length(alignment, sequence):
"""Trim alignment to a sequence, i.e. leave only postions that correspond to this sequence span"""
n1 = str(uuid.uuid4())
n2 = str(uuid.uuid4())
#Get consensus
a = SummaryInfo(alignment)
cons = a.dumb_consensus(threshold=0.1, ambiguous='X')
#Needle it
SeqIO.write([SeqRecord(cons, id='CONS', name='CONS')], n1 + '.fasta',
'fasta')
SeqIO.write([SeqRecord(sequence, id='KEY', name='KEY')], n2 + '.fasta',
'fasta')
#Now we will redo it with Needlman Wunsh - the global alignment
needle_cline = NeedleCommandline(asequence=n1 + ".fasta",
bsequence=n2 + ".fasta",
gapopen=10,
gapextend=0.5,
outfile=n1 + ".txt")
stdout, stderr = needle_cline()
# print('Needle alignment')
align = AlignIO.read(n1 + ".txt", "emboss")
os.system('rm %s.fasta %s.fasta %s.txt' % (n1, n2, n1))
# print align
# print alignment
#first seq is consensus, we need to get borders useing second one.
seq = str(align[1, :].seq)
# print seq
begin = seq.index(str(sequence[0]))
end = len(seq) - seq[::-1].index(str(sequence[-1]))
print begin
print end
return alignment[:, begin:end]
def test_nucleotides(self):
filename = "GFF/multi.fna"
format = "fasta"
alignment = AlignIO.read(filename, format, alphabet=unambiguous_dna)
summary = SummaryInfo(alignment)
c = summary.dumb_consensus(ambiguous="N")
self.assertEqual(str(c), "NNNNNNNN")
c = summary.gap_consensus(ambiguous="N")
self.assertEqual(str(c), "NNNNNNNN")
expected = {"A": 0.25, "G": 0.25, "T": 0.25, "C": 0.25}
m = summary.pos_specific_score_matrix(chars_to_ignore=["-"],
axis_seq=c)
self.assertEqual(
str(m), """ A C G T
N 2.0 0.0 1.0 0.0
N 1.0 1.0 1.0 0.0
N 1.0 0.0 2.0 0.0
N 0.0 1.0 1.0 1.0
N 1.0 2.0 0.0 0.0
N 0.0 2.0 1.0 0.0
N 1.0 2.0 0.0 0.0
N 0.0 2.0 1.0 0.0
""")
# Have a generic alphabet, without a declared gap char, so must tell
# provide the frequencies and chars to ignore explicitly.
ic = summary.information_content(e_freq_table=expected,
chars_to_ignore=["-"])
self.assertAlmostEqual(ic, 7.32029999423075, places=6)
def _run_hmmer_single_thread(sequences, iter=3):
pbar = tqdm(range(len(sequences)), desc="sequences processed")
database = "data/Uniprot/uniprot_sprot.fasta"
outpth = os.path.join(tmp_dir, "align.sto")
for seqid, seq in sequences.items():
pbar.update(1)
cline = "jackhmmer -N %d --acc --noali -A %s - %s" % (iter, outpth, database)
child = subprocess.Popen(cline,
stdin=subprocess.PIPE,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
universal_newlines=True,
shell=(sys.platform != "win32"))
stdout, _ = child.communicate(input=">%s\n%s" % (seqid, seq))
assert child.returncode == 0
info = SummaryInfo(AlignIO.read(outpth, "stockholm"))
pssm = info.pos_specific_score_matrix(chars_to_ignore=[GAP])
db.pssm.update_one({
"_id": seqid}, {
'$set': {"pssm": pssm.pssm, "seq": seq}
}, upsert=True)
pbar.close()
def get_hist_ss_in_aln_as_string(alignment, type='Unknown', debug=0):
""" gets an annotation line as a string, borrowed from aln2html"""
sinfo = SummaryInfo(alignment)
cons = sinfo.gap_consensus(threshold=0.9, ambiguous='X')
features = get_hist_ss_in_aln_for_html(alignment, type=type, debug=0)
f_description = ''
annot_line = [0, 1, 2]
if (features):
annot_line[0] = list(' ' * len(cons))
annot_line[1] = list(' ' * len(cons))
annot_line[2] = list(' ' * len(cons))
keys = sorted(list(features.keys()), key=lambda x: x[0])
for k in keys:
if (features[k].get('description', 0)):
f_description += '{0}-{1};'.format(features[k]['symbol'],
features[k]['description'])
lev = features[k].get('level', 0)
if (re.match('^\s+$', ''.join(annot_line[lev][k[0]:k[1] + 1]))):
annot_line[lev][k[0]:k[1] +
1] = features[k]['symbol'] * (k[1] - k[0] + 1)
else:
lev += 1
if (re.match('^\s+$', ''.join(annot_line[lev][k[0]:k[1] + 1]))):
annot_line[lev][k[0]:k[1] +
1] = features[k]['symbol'] * (k[1] - k[0] + 1)
else:
lev += 1
if (re.match('^\s+$',
''.join(annot_line[lev][k[0]:k[1] + 1]))):
annot_line[lev][k[0]:k[1] +
1] = features[k]['symbol'] * (k[1] - k[0] +
1)
return annot_line[0] #other are ignored currently
def get_features_in_aln(alignment, variant, save_dir="", save_gff=True):
#Let's extract consensus
a = SummaryInfo(alignment)
cons = a.dumb_consensus(threshold=0.1, ambiguous='X')
seq = Sequence(id="Consensus",
variant_id=variant,
taxonomy_id=1,
sequence=str(cons))
updated_features = get_variant_features(seq,
save_dir=save_dir,
save_gff=save_gff)
return updated_features
def trim_aln_gaps(alignment, threshold=0.8):
"""Removes positions with more than threshold gaps in alignment"""
a = SummaryInfo(alignment)
cons = a.gap_consensus(threshold=threshold, ambiguous='X')
new_aln = alignment[:, 0:0]
for c, i in zip(cons, range(len(cons))):
if (c == '-'):
continue
else:
new_aln += alignment[:, i:i + 1]
return new_aln
def get_hist_ss_in_aln(alignment, type='Unknown', debug=0):
"""Returns sequence elements in histone alignment, all numbers assume first element in seq has number 0!!! Not like in PDB"""
#Let's extract consensus
if (debug):
print(alignment)
a = SummaryInfo(alignment)
cons = a.gap_consensus(threshold=0.5, ambiguous='X')
cons = Seq(str(cons).replace('-', 'X'))
if (debug):
print("Consensus")
print(cons)
hv, ss = get_hist_ss(cons, type, debug)
return hv, ss
def make_consensus(alignment_file, format_alignment):
from Bio import AlignIO
from Bio.Align.AlignInfo import SummaryInfo
format = format_alignment
alignment = AlignIO.read(alignment_file, format)
summary = SummaryInfo(alignment)
mat = summary.pos_specific_score_matrix()
consensus = ""
l = list()
for row in mat:
key, freq = max(row.items(), key=operator.itemgetter(1))
consensus = "".join([consensus, key])
l.append(freq / len(alignment._records))
consensus = consensus.strip("-")
return consensus, l
def test_iupac_consensus(self):
filename = "GFF/multi_1.fna"
format = "fasta"
alignment = AlignIO.read(filename, format, alphabet=unambiguous_dna)
summary = SummaryInfo(alignment)
c = summary.iupac_consensus(ambiguous="N")
self.assertEqual(str(c), 'TAWSRWYSNBATCG')
self.assertNotEqual(c.alphabet, unambiguous_dna)
self.assertTrue(isinstance(c.alphabet, DNAAlphabet))
c = summary.gap_iupac_consensus(ambiguous="N")
self.assertEqual(str(c), 'TAWSRWYSNBATCG')
self.assertNotEqual(c.alphabet, unambiguous_dna)
self.assertTrue(isinstance(c.alphabet, DNAAlphabet))
def getConservedDomain(self):
cons = []
align = SummaryInfo(self.__alignment)
consenso = str(align.gap_consensus())
temp = ''
for i in range(len(consenso)):
if consenso[i] not in "X-":
temp += consenso[i]
else:
if temp != '':
cons.append(temp)
temp = ''
max_cons = ''
for i in cons:
if len(i) > len(max_cons): max_cons = i
return max_cons
def get_seed_aln_and_features(request, seed):
from Bio.Align import MultipleSeqAlignment
from Bio.Align.AlignInfo import SummaryInfo
seed_file = os.path.join(settings.STATIC_ROOT_AUX, "browse", "seeds")
try:
histone = Histone.objects.get(id=seed)
seed_file = os.path.join(seed_file, "{}".format(histone.id))
except Histone.DoesNotExist:
try:
variant = Variant.objects.get(id=seed)
seed_file = os.path.join(seed_file, variant.hist_type.id, "{}".format(variant.id))
# the default names for canonical are with underscores, so we do not need to convert back. ALEXEY, 30/12/15
# seed_file = os.path.join(seed_file, variant.hist_type.id, "{}".format(variant.id.replace("canonical", "canonical_")))
except Variant.DoesNotExist:
return HttpResponseNotFound('<h1>No histone variant with name {}</h1>'.format(seed))
download = request.GET.get("download", False) == "true"
format = request.GET.get("format", "json")
try:
limit = int(request.GET.get("limit", 0))
except ValueError:
limit = 0
consensus = request.GET.get("consensus", False)
if not consensus in ["limit", "all", False]:
consensus = "all"
response = HttpResponse(content_type='text')
if download:
response['Content-Disposition'] = 'attachment; filename="{}.{}"'.format(seed, format)
sequences = SeqIO.parse("{}.fasta".format(seed_file), "fasta")
if consensus:
sequences = [s for i, s in enumerate(sequences) if consensus == "all" or (consensus == "limit" and i < limit)]
msa = MultipleSeqAlignment(sequences)
a = SummaryInfo(msa)
sequences.insert(0, SeqRecord(id="Consensus", description="", seq=a.dumb_consensus(threshold=0.1, ambiguous='X')))
limit = limit+1 if limit > 0 else 0
def limited_seqs():
for i, seq in enumerate(sequences):
if not consensus or consensus == "limit" or (limit > 0 and i < limit):
yield seq
with open("{}.gff".format(seed_file)) as feature_file:
features = feature_file.read()
if format == "fasta":
SeqIO.write(limited_seqs(), response, "fasta")
elif format == "gff":
response.write(features)
elif format == "pdf":
with open("{}.pdf".format(seed_file)) as pdf:
response.write(pdf.read())
else:
#Default format is json
sequences = [{"name":seq.id, "seq":seq.seq.tostring()} for seq in limited_seqs()]
result = {"seqs":sequences, "features":features}
response.write(json.dumps(result))
return response
def aln2html(msa,filename,features=None,title=None,description=True,field1w=20,field2w=35):
"""
This function outputs HTML from msa and annotates features.
msa - Biopython MSA,
filename - html file to output the result.
features - a dictionary of features, organized as follows:
{(begin,end):{'level':0(default),'symbol':'H','description':'desc'}}
if features overlap and not levels, they will be split to different levels.
Only three levels (0,1,2) are available.
"""
style="""
pre,td{margin: 0px;padding: 0px;border: 0px;}
.pos{color:blue;}
.neg{color:red;}
.pol{color:green;}
.hphob{color:grey;}
.def{color:black;}
.conserved{background:lightblue;}
.nonconserved{background:white;}
"""
sinfo=SummaryInfo(msa)
cons=sinfo.gap_consensus(threshold=0.9, ambiguous='X')
#Let's work on features
f_description=''
msatext=''
annot_line=[0,1,2]
if(features):
annot_line[0]=list(' '*len(cons))
annot_line[1]=list(' '*len(cons))
annot_line[2]=list(' '*len(cons))
keys=sorted(list(features.keys()),key=lambda x: x[0])
for k in keys:
if(features[k].get('description',0)):
f_description+='{0}-{1};'.format(features[k]['symbol'],features[k]['description'])
lev=features[k].get('level',0)
if(re.match('^\s+$',''.join(annot_line[lev][k[0]:k[1]+1]))):
annot_line[lev][k[0]:k[1]+1]=features[k]['symbol']*(k[1]-k[0]+1)
else:
lev+=1
if(re.match('^\s+$',''.join(annot_line[lev][k[0]:k[1]+1]))):
annot_line[lev][k[0]:k[1]+1]=features[k]['symbol']*(k[1]-k[0]+1)
else:
lev+=1
if(re.match('^\s+$',''.join(annot_line[lev][k[0]:k[1]+1]))):
annot_line[lev][k[0]:k[1]+1]=features[k]['symbol']*(k[1]-k[0]+1)
if(not re.match('^\s+$',''.join(annot_line[2]))):
msatext='<TR><TD><PRE>{0:<{field1w}}</PRE></TD>'.format('annotation',field1w=field1w+2)
if(description):
msatext+='<TD><PRE>{0:<{field2w}}</PRE></TD>'.format('level 2',field2w=field2w+2)
for c in annot_line[2]:
msatext+='<TD><PRE>{0}</PRE></TD>'.format(c)
msatext+='</TR>'
if(not re.match('^\s+$',''.join(annot_line[1]))):
msatext='<TR><TD><PRE>{0:<{field1w}}</PRE></TD>'.format('annotation',field1w=field1w+2)
if(description):
msatext+='<TD><PRE>{0:<{field2w}}</PRE></TD>'.format('level 1',field2w=field2w+2)
for c in annot_line[1]:
msatext+='<TD><PRE>{0}</PRE></TD>'.format(c)
msatext+='</TR>'
if(not re.match('^\s+$',''.join(annot_line[0]))):
msatext='<TR><TD><PRE>{0:<{field1w}}</PRE></TD>'.format('annotation',field1w=field1w+2)
if(description):
msatext+='<TD><PRE>{0:<{field2w}}</PRE></TD>'.format('level 0',field2w=field2w+20)
for c in annot_line[0]:
msatext+='<TD><PRE>{0}</PRE></TD>'.format(c)
msatext+='</TR>'
f_description+='<BR><BR>'
for s in msa:
if(re.search(r'\d\d+',s.id)):
gi=re.search(r'(\d\d+)',s.id).group(1)
line='<TR><TD><PRE><a href="http://www.ncbi.nlm.nih.gov/protein/?term={0}">{1:<{field1w}}</a></PRE></TD>'.format(gi,s.id[:field1w],field1w=field1w+2)
if(description):
line+='<TD><PRE>{0:<{field2w}}</PRE></TD>'.format(s.description[:field2w],field2w=field2w+2)
else:
line='<TR><TD><PRE>{0:<{field1w}}</PRE></TD>'.format(s.id[:field1w],field1w=field1w+2)
if(description):
line+='<TD><PRE>{0:<{field2w}}</PRE></TD>'.format(s.description[:field2w],field1w=field1w+2)
for c,i in zip(s.seq,range(len(s.seq))):
line+='<TD><PRE class="{0} {1}">{2}</PRE></TD>'.format(restypedict.get(c,'def'),'conserved' if c==cons[i] and c!='-' else 'nonconserved',c)
line+='</TR>'
msatext=msatext+line
a=open(filename,'w')
a.write("""
<!DOCTYPE html>
<HTML>
<HEAD>
<META http-equiv="Content-Type" content="text/html; charset=utf-8"/>
<TITLE>MultipleSequenceAlignment</TITLE>
<style>
{style}
</style>
</HEAD>
<BODY style="background-color:white; color:black; a:link:blue; a:active:red; a:visited:purple">
{title}<BR><BR>
{features}
<TABLE style="border:0px; border-spacing:0px; background-color:white; color:black; a:link:blue; a:active:red; a:visited:purple;">
{msatext}
</TABLE>
</BODY>
</HTML>
""".format(\
title=title,\
msatext=msatext,\
style=style,\
features=f_description
))
a.close()
def get_aln_and_features(request, ids=None):
from tools.hist_ss import get_variant_features
from tools.L_shade_hist_aln import write_alignments
import subprocess
import StringIO
from Bio.Align import MultipleSeqAlignment
from Bio.Align.AlignInfo import SummaryInfo
from Bio.SeqRecord import SeqRecord
save_dir = os.path.join(os.path.sep, "tmp", "HistoneDB")
if not os.path.exists(save_dir):
os.makedirs(save_dir)
os.chmod(save_dir,0o777)
if ids is None and request.method == "GET" and "id" in request.GET:
ids = request.GET.getlist("id")
sequences = Sequence.objects.filter(id__in=ids)
download = False
upload = False
elif request.GET.get("download", False) == "true":
download = True
upload = False
else:
#Returning 'false' stops Bootstrap table
return "false"
if request.GET.get("upload", False) == "true":
uploaded_sequence = request.session.get("uploaded_sequences", [])
if len(uploaded_sequence) > 0:
try:
variant = Variant.objects.get(id=uploaded_sequence[0]["variant"])
except:
if len(sequences) > 0:
variant = sequences[0].variant
else:
return "false"
uploaded_sequence = Sequence(
id=uploaded_sequence[0]["id"],
variant=variant,
sequence=uploaded_sequence[0]["sequence"],
taxonomy=Taxonomy.objects.get(name=uploaded_sequence[0]["taxonomy"]))
upload = True
download = False
if not download:
if len(sequences) == 0:
return None, None
elif len(sequences) == 1:
#Already aligned to core histone
seq = sequences[0]
hist_type = seq.variant.hist_type.id
variants = [seq.variant]
if upload:
sequences = [uploaded_sequence, seq]
else:
#let's load the corresponding canonical
try:
if(("canonical" in str(seq.variant)) or ("generic" in str(seq.variant))):
canonical=seq
elif(str(seq.variant.hist_type)=="H1"):
canonical=Sequence.objects.filter(variant_id='generic_'+str(seq.variant.hist_type),reviewed=True,taxonomy=seq.taxonomy)[0]
else:
canonical=Sequence.objects.filter(variant_id='canonical_'+str(seq.variant.hist_type),reviewed=True,taxonomy=seq.taxonomy)[0]
except:
try: #try H2A.X as a substitute for canonical
if(str(seq.variant.hist_type)=='H2A'):
canonical=Sequence.objects.filter(variant_id='H2A.X',reviewed=True,taxonomy=seq.taxonomy)[0]
elif(str(seq.variant.hist_type)=='H3'): #Try H3.3
canonical=Sequence.objects.filter(variant_id='H3.3',reviewed=True,taxonomy=seq.taxonomy)[0]
elif(str(seq.variant.id)=='scH1'):
canonical=seq
else:
raise
except:
canonical=seq #we here default not to show the sequence by simply suppling itslef - only one line will be displayed
#default Xenopus
# if(str(seq.variant.hist_type)=="H1"):
# canonical = Sequence(id="0000|xenopus|generic{}".format(hist_type), sequence=str(TemplateSequence.objects.get(variant="General{}".format(hist_type)).get_sequence().seq))
# else:
# canonical = Sequence(id="0000|xenopus|canonical{}".format(hist_type), sequence=str(TemplateSequence.objects.get(variant="General{}".format(hist_type)).get_sequence().seq))
sequences = [canonical, seq]
sequence_label = seq.short_description
else:
seq = sequences[0]
variants = list(Variant.objects.filter(id__in=sequences.values_list("variant", flat=True).distinct()))
sequence_label = "Consensus"
muscle = os.path.join(os.path.dirname(sys.executable), "muscle")
process = subprocess.Popen([muscle], stdin=subprocess.PIPE, stdout=subprocess.PIPE)
sequences = "\n".join([s.format() for s in sequences])
aln, error = process.communicate(sequences)
seqFile = StringIO.StringIO()
seqFile.write(aln)
seqFile.seek(0)
sequences = list(SeqIO.parse(seqFile, "fasta")) #Not in same order, but does it matter?
msa = MultipleSeqAlignment(sequences)
a = SummaryInfo(msa)
cons = Sequence(id=sequence_label, variant_id=variants[0].id, taxonomy_id=1, sequence=a.dumb_consensus(threshold=0.1, ambiguous='X').tostring())
save_dir = os.path.join(os.path.sep, "tmp", "HistoneDB")
if not os.path.exists(save_dir):
os.makedirs(save_dir)
features = get_variant_features(cons, variants=variants, save_dir=save_dir)
#A hack to avoid two canonical seqs
unique_sequences = [sequences[0]] if len(sequences) == 2 and sequences[0].id == sequences[1].id else sequences
# doing the Sequence.short_description work
#Note that the gffs are also generated with the short description not
sequences = [{"name":"QUERY" if "QUERY" in s.id else Sequence.long_to_short_description(s.id), "seq":s.seq.tostring()} for s in unique_sequences]
# sequences = [{"name":s.id, "seq":s.seq.tostring()} for s in sequences]
if sequence_label == "Consensus":
sequences.insert(0, cons.to_dict(id=True))
request.session["calculated_msa_seqs"] = sequences
request.session["calculated_msa_features"] = features#.to_ict() if features else {}
result = {"seqs":sequences, "features":features} #.full_gff() if features else ""}
return JsonResponse(result, safe=False)
else:
format = request.GET.get("format", "json")
response = HttpResponse(content_type='text')
response['Content-Disposition'] = 'attachment; filename="sequences.{}"'.format(format)
sequences = request.session.get("calculated_msa_seqs", [])
features = request.session.get("calculated_msa_features", "")
#features = Features.from_dict(Sequence("Consensus"), features_dict) if features_dict else None
if format == "fasta":
for s in sequences:
print >> response, ">{}\n{}".format(s["name"], s["seq"])
elif format == "gff":
response.write(features) #.full_gff() if features else "")
elif format == "pdf":
aln = MultipleSeqAlignment([SeqRecord(Seq(s["seq"]), id=s["name"]) for s in sequences[1:]])
result_pdf = write_alignments(
[aln],
save_dir = save_dir
)
with open(result_pdf) as pdf:
response.write(pdf.read())
#Cleanup
os.remove(result_pdf)
else:
#Default format is json
result = {"seqs":sequences, "features":features} #.full_gff() if features else ""}
response.write(json.dumps(result))
return response
def getConsenso(self):
align = SummaryInfo(self.__alignment)
return align.gap_consensus()
def getPerc(self):
align = SummaryInfo(self.__alignment)
return float(1 - (align.gap_consensus().count("X") +
align.gap_consensus().count("-")) /
len(str(align.gap_consensus())))