def get_protein_seq(self, transcript_id: str):
"""
Extract amino acid sequence for given transcript_id
:param transcript_id:
:return: amino acid sequence
"""
return translate(self.get_seq(transcript_id), hg38=True)
def _prepare_seq(cls, *args, **kwargs):
"""
Prepare the dna sequence and translate it into amino acid sequence
:param seqs: current dna sequence
:param intervals: the list of intervals corresponding to the sequence snippets
:param reverse_complement: should the dna be reverse-complemented?
:return: amino acid sequence
"""
return translate(super()._prepare_seq(*args, **kwargs), hg38=True)
def test_ensembl_uniprot_seq(tse):
id_and_seq = {}
with open(uniprot_seq_ref, 'r+') as f:
key = ""
for line in f:
if '>' in line:
key = (line.replace('>', '')).rstrip()
else:
id_and_seq[key] = line.rstrip()
for transkript_id, ref_seq in tqdm(id_and_seq.items()):
test_seq = translate(tse.get_seq(transkript_id), True)
assert test_seq == ref_seq, test_seq
def test_vcf_single_variant_synonymous_mutations(tse, svp):
transcript_id = 'ENST00000356175'
ref_seq = translate(tse.get_seq(transcript_id), True)
single_var_seq = list(svp.extract(transcript_id))
for seq in single_var_seq:
assert seq == ref_seq, seq
assert len(
single_var_seq) == 337, 'Number of sequences != number of variants'
count = 0
single_var_seq = list(svp.extract_all())
for t_id in single_var_seq:
count += len(list(t_id))
assert count == 825
def test_hg38(tse):
with open('err_transcripts', 'w+') as f:
dfp = read_pep_fa(protein_file)
dfp['transcript_id'] = dfp.transcript.str.split(".", n=1,
expand=True)[0]
#assert not dfp['transcript_id'].duplicated().any()
dfp = dfp.set_index("transcript_id")
#dfp = dfp[~dfp.chromosome.isnull()]
assert len(tse) > 100
assert tse.transcripts.isin(dfp.index).all()
div3_error = 0
seq_mismatch_err = 0
err_transcripts = []
for transcript_id in tqdm(tse.transcripts):
# make sure all ids can be found in the proteome
dna_seq = tse.get_seq(transcript_id)
if dna_seq == "NNN":
f.write(transcript_id +
' has an ambiguous start and end.Skip!')
continue
# dna_seq = dna_seq[:(len(dna_seq) // 3) * 3]
# if len(dna_seq) % 3 != 0:
# div3_error += 1
# print("len(dna_seq) % 3 != 0: {}".format(transcript_id))
# err_transcripts.append({"transcript_id": transcript_id, "div3_err": True})
# continue
if len(dna_seq) % 3 != 0:
f.write(transcript_id)
continue
prot_seq = translate(dna_seq, hg38=True)
if dfp.loc[transcript_id].seq != prot_seq:
seq_mismatch_err += 1
f.write("seq.mismatch: {}".format(transcript_id))
n_mismatch = 0
for i in range(len(prot_seq)):
a = dfp.loc[transcript_id].seq[i]
b = prot_seq[i]
if a != b:
n_mismatch += 1
f.write("{} {} {}/{}".format(a, b, i, len(prot_seq)))