def main(input, blastndb, output, probe_length, search_step, evalue,
blastn_tmpdir, threads):
"""
Search unique mapped probe(sub-sequence)
within a series of sequences stored in a fasta file.
\b
For example:
select 30 candidate probe regions with length 500bp, firstly,
```
$ python uniformly_spaced.py data/hg19.fa ./candidate.fa chr1:89000000-90000000 -n 30 -l 500
```
then select unique maped probe(sub-sequence) from it.
```
$ python search_uniq.py candidate.fa example/blastn_db/hg19 probe.fa
```
\b
Args
----
input : str
Path to input fasta file.
blastndb : str
Path to blastn database.
build with `makeblastdb` command.
output : str
Path to output fasta file.
"""
with open(input) as f:
input_seqs = FastaIO.FastaIterator(f)
probes = search_passed_probes(input_seqs, blastndb, evalue,
probe_length, search_step, blastn_tmpdir,
threads)
save_fasta(probes, output)
def load_data(k, stride, pos_fasta, neg_fasta):
vocab = Vocabulary(k=k)
X = []
n_pos = 0
n_neg = 0
for fasta in pos_fasta, neg_fasta:
with open(fasta) as f:
for s in tqdm(FastaIO.FastaIterator(f)):
seq = str(s.seq)
if vocab.unknow_char in seq:
continue
try:
x = vocab.kmer_count(seq, stride)
except AssertionError:
continue
X.append(x)
if fasta == pos_fasta:
n_pos += 1
else:
n_neg += 1
X = np.vstack(X)
y = np.hstack([np.ones(n_pos), np.zeros(n_neg)])
return X, y
def input_text_to_df(input_text):
"""Converts fasta contents to a df with columns sequence_name and sequence."""
with io.StringIO(initial_value=input_text) as f:
fasta_records = list(FastaIO.FastaIterator(f))
fasta_df = pd.DataFrame([(f.name, str(f.seq)) for f in fasta_records],
columns=['sequence_name', 'sequence'])
return fasta_df
def _assert_fasta_parsable(input_text):
with io.StringIO(initial_value=input_text) as f:
fasta_itr = FastaIO.FastaIterator(f)
end_iteration_sentinel = object()
# Avoid parsing the entire FASTA contents by using `next`.
# A malformed FASTA file will have no entries in its FastaIterator.
# This is unfortunate (instead of it throwing an error).
if next(fasta_itr, end_iteration_sentinel) is end_iteration_sentinel:
raise ValueError(
'Failed to parse any input from fasta file. '
'Consider checking the formatting of your fasta file. '
'First bit of contents from the fasta file was\n'
'{}'.format(input_text.splitlines()[:3]))
help="An ensembl cdna fasta file with the properly formatted header")
parser.add_argument("--output",
default=None,
type=str,
help="output tr2gene file")
args = parser.parse_args()
L.info("args:")
print(args)
outf = iotools.open_file(args.output, "w")
with iotools.open_file(args.fasta, "r") as handle:
for record in FastaIO.FastaIterator(handle):
description = record.description
trans = description.split(" ")[0]
m = re.search('gene:(\S+)', description)
gene = m.group(1)
try:
x = re.search('gene_symbol:(\S+)', description)
symbol = x.group(1)
except Exception:
pass
outf.write("%s\t%s\t%s\n" % (trans, gene, symbol))
outf.close()
