def _alt_ac_for_tx_ac(self, tx_ac):
"""return chromosomal accession for given transcript accession (and
the_assembly and aln_method setting used to instantiate this
AssemblyMapper)
"""
alt_acs = [
e["alt_ac"] for e in self.hdp.get_tx_mapping_options(tx_ac)
if e["alt_aln_method"] == self.alt_aln_method
and e["alt_ac"] in self._assembly_accessions
]
if not alt_acs:
raise HGVSDataNotAvailableError(
"No alignments for {tx_ac} in {an} using {am}".format(
tx_ac=tx_ac, an=self.assembly_name,
am=self.alt_aln_method))
# TODO: conditional is unnecessary; remove
if len(alt_acs) > 1:
names = set(self._assembly_map[ac] for ac in alt_acs)
if names != set("XY"):
alts = ", ".join([
"{ac} ({n})".format(ac=ac, n=self._assembly_map[ac])
for ac in alt_acs
])
raise HGVSError(
"Multiple chromosomal alignments for {tx_ac} in {an}"
" using {am} (non-pseudoautosomal region) [{alts}]".format(
tx_ac=tx_ac,
an=self.assembly_name,
am=self.alt_aln_method,
alts=alts))
# assume PAR
if self.in_par_assume is None:
raise HGVSError(
"Multiple chromosomal alignments for {tx_ac} in {an}"
" using {am} (likely pseudoautosomal region)".format(
tx_ac=tx_ac,
an=self.assembly_name,
am=self.alt_aln_method))
alt_acs = [
ac for ac in alt_acs
if self._assembly_map[ac] == self.in_par_assume
]
if len(alt_acs) != 1:
raise HGVSError(
"Multiple chromosomal alignments for {tx_ac} in {an}"
" using {am}; in_par_assume={ipa} selected {n} of them".
format(tx_ac=tx_ac,
an=self.assembly_name,
am=self.alt_aln_method,
ipa=self.in_par_assume,
n=len(alt_acs)))
assert len(
alt_acs) == 1, "Should have exactly one alignment at this point"
return alt_acs[0]
def __str__(self):
if self.min > self.max:
raise HGVSError(
"Repeat min count must be less than or equal to max count")
if self.min == self.max:
return "{self.ref}[{self.min}]".format(self=self)
return "{self.ref}({self.min}_{self.max})".format(self=self)
def c_to_n(self, c_interval):
"""convert a transcript CDS (c.) interval to a transcript cDNA (n.) interval"""
if self.cds_start_i is None: # cds_start_i defined iff cds_end_i defined; see assertion above
raise HGVSUsageError(
"CDS is undefined for {self.tx_ac}; cannot map from c. coordinate (non-coding transcript?)".format(
self=self))
# start
if c_interval.start.datum == hgvs.location.CDS_START and c_interval.start.base < 0:
rs = c_interval.start.base + self.cds_start_i + 1
elif c_interval.start.datum == hgvs.location.CDS_START and c_interval.start.base > 0:
rs = c_interval.start.base + self.cds_start_i
elif c_interval.start.datum == hgvs.location.CDS_END:
rs = c_interval.start.base + self.cds_end_i
# end
if c_interval.end.datum == hgvs.location.CDS_START and c_interval.end.base < 0:
re = c_interval.end.base + self.cds_start_i + 1
elif c_interval.end.datum == hgvs.location.CDS_START and c_interval.end.base > 0:
re = c_interval.end.base + self.cds_start_i
elif c_interval.end.datum == hgvs.location.CDS_END:
re = c_interval.end.base + self.cds_end_i
if rs <= 0 or re > self.tgt_len:
raise HGVSError("The given coordinate is outside the bounds of the reference sequence.")
n_interval = hgvs.location.BaseOffsetInterval(
start=hgvs.location.BaseOffsetPosition(
base=rs, offset=c_interval.start.offset, datum=hgvs.location.SEQ_START),
end=hgvs.location.BaseOffsetPosition(base=re, offset=c_interval.end.offset, datum=hgvs.location.SEQ_START),
uncertain=c_interval.uncertain)
return n_interval
def format(self, conf=None):
if self.ref is None and self.alt is None:
raise HGVSError("RefAlt: ref and alt sequences are both undefined")
max_ref_length = self._format_config_na(conf)
if max_ref_length is not None:
ref = self.ref_s
if ref is None or len(ref) > max_ref_length:
ref = ''
else:
ref = self.ref
# subst and delins
if self.ref is not None and self.alt is not None:
if self.ref == self.alt:
s = "{ref}=".format(ref=ref)
elif len(self.alt) == 1 and len(
self.ref
) == 1 and not self.ref.isdigit(): # don't turn del5insT into 5>T
s = "{self.ref}>{self.alt}".format(self=self)
else:
s = "del{ref}ins{alt}".format(ref=ref, alt=self.alt)
# del case
elif self.ref is not None:
s = "del{ref}".format(ref=ref)
# ins case
else: # self.alt is not None
s = "ins{self.alt}".format(self=self)
return "(" + s + ")" if self.uncertain else s
def is_coding_transcript(self):
if ((self.tx_info["cds_start_i"] is not None) ^
(self.tx_info["cds_end_i"] is not None)):
raise HGVSError(
"{self.tx_ac}: CDS start_i and end_i"
" must be both defined or both undefined".format(self=self))
return self.tx_info["cds_start_i"] is not None
def _get_cursor(self, n_retries=1):
"""Returns a context manager for obtained from a single or pooled
connection, and sets the PostgreSQL search_path to the schema
specified in the connection URL.
Although *connections* are threadsafe, *cursors* are bound to
connections and are *not* threadsafe. Do not share cursors
across threads.
Use this funciton like this::
with hdp._get_cursor() as cur:
# your code
Do not call this function outside a contextmanager.
"""
n_tries_rem = n_retries + 1
while n_tries_rem > 0:
try:
conn = self._pool.getconn() if self.pooling else self._conn
# autocommit=True obviates closing explicitly
conn.autocommit = True
cur = conn.cursor(cursor_factory=psycopg2.extras.DictCursor)
cur.execute(
"set search_path = {self.url.schema};".format(self=self))
yield cur
# contextmanager executes these when context exits
cur.close()
if self.pooling:
self._pool.putconn(conn)
break
except psycopg2.OperationalError:
_logger.warning(
"Lost connection to {url}; attempting reconnect".format(
url=self.url))
if self.pooling:
self._pool.closeall()
self._connect()
_logger.warning("Reconnected to {url}".format(url=self.url))
n_tries_rem -= 1
else:
# N.B. Probably never reached
raise HGVSError(
"Permanently lost connection to {url} ({n} retries)".format(
url=self.url, n=n_retries))
def __init__(self, hdp, tx_ac, alt_ac, alt_aln_method):
self.hdp = hdp
self.tx_ac = tx_ac
self.alt_ac = alt_ac
self.alt_aln_method = alt_aln_method
if self.alt_aln_method != "transcript":
self.tx_info = hdp.get_tx_info(self.tx_ac, self.alt_ac,
self.alt_aln_method)
if self.tx_info is None:
raise HGVSDataNotAvailableError(
"TranscriptMapper(tx_ac={self.tx_ac}, "
"alt_ac={self.alt_ac}, alt_aln_method={self.alt_aln_method}): "
"No transcript info".format(self=self))
self.tx_exons = hdp.get_tx_exons(self.tx_ac, self.alt_ac,
self.alt_aln_method)
if self.tx_exons is None:
raise HGVSDataNotAvailableError(
"TranscriptMapper(tx_ac={self.tx_ac}, "
"alt_ac={self.alt_ac}, alt_aln_method={self.alt_aln_method}): "
"No transcript exons".format(self=self))
# hgvs-386: An assumption when building the cigar string
# is that exons are adjacent. Assert that here.
tx_exons = sorted(self.tx_exons, key=lambda e: e["ord"])
for i in range(1, len(tx_exons)):
if tx_exons[i - 1]["tx_end_i"] != tx_exons[i]["tx_start_i"]:
raise HGVSDataNotAvailableError(
"TranscriptMapper(tx_ac={self.tx_ac}, "
"alt_ac={self.alt_ac}, alt_aln_method={self.alt_aln_method}): "
"Exons {a} and {b} are not adjacent".format(self=self,
a=i,
b=i + 1))
self.strand = self.tx_exons[0]["alt_strand"]
self.cds_start_i = self.tx_info["cds_start_i"]
self.cds_end_i = self.tx_info["cds_end_i"]
self.gc_offset = self.tx_exons[0]["alt_start_i"]
self.cigar = build_tx_cigar(self.tx_exons, self.strand)
self.im = hgvs.intervalmapper.IntervalMapper.from_cigar(self.cigar)
self.tgt_len = self.im.tgt_len
else:
# this covers the identity cases n <-> c
self.tx_identity_info = hdp.get_tx_identity_info(self.tx_ac)
if self.tx_identity_info is None:
raise HGVSError(
"TranscriptMapper(tx_ac={self.tx_ac}, "
"alt_ac={self.alt_ac}, alt_aln_method={self.alt_aln_method}): "
"No transcript identity info".format(self=self))
self.cds_start_i = self.tx_identity_info["cds_start_i"]
self.cds_end_i = self.tx_identity_info["cds_end_i"]
self.tgt_len = sum(self.tx_identity_info["lengths"])
assert not (
(self.cds_start_i is None) ^ (self.cds_end_i is None)
), "CDS start and end must both be defined or neither defined"
def format(self, conf=None):
if self.min > self.max:
raise HGVSError("Repeat min count must be less than or equal to max count")
max_ref_length = self._format_config_na(conf)
ref = self.ref
if max_ref_length is not None and (ref is None or len(ref) > max_ref_length):
ref = ''
if self.min == self.max:
return "{ref}[{min}]".format(ref=ref, min=self.min)
return "{ref}({min}_{max})".format(ref=ref, min=self.min, max=self.max)
def n_to_c(self, n_interval):
"""convert a transcript cDNA (n.) interval to a transcript CDS (c.) interval"""
if self.cds_start_i is None: # cds_start_i defined iff cds_end_i defined; see assertion above
raise HGVSUsageError(
"CDS is undefined for {self.tx_ac}; cannot map to c. coordinate (non-coding transcript?)"
.format(self=self))
if n_interval.start.base <= 0 or n_interval.end.base > self.tgt_len:
raise HGVSError(
"The given coordinate is outside the bounds of the reference sequence."
)
# start
if n_interval.start.base <= self.cds_start_i:
cs = n_interval.start.base - (self.cds_start_i + 1)
cs_datum = Datum.CDS_START
elif n_interval.start.base > self.cds_start_i and n_interval.start.base <= self.cds_end_i:
cs = n_interval.start.base - self.cds_start_i
cs_datum = Datum.CDS_START
else:
cs = n_interval.start.base - self.cds_end_i
cs_datum = Datum.CDS_END
# end
if n_interval.end.base <= self.cds_start_i:
ce = n_interval.end.base - (self.cds_start_i + 1)
ce_datum = Datum.CDS_START
elif n_interval.end.base > self.cds_start_i and n_interval.end.base <= self.cds_end_i:
ce = n_interval.end.base - self.cds_start_i
ce_datum = Datum.CDS_START
else:
ce = n_interval.end.base - self.cds_end_i
ce_datum = Datum.CDS_END
c_interval = hgvs.location.BaseOffsetInterval(
start=hgvs.location.BaseOffsetPosition(
base=cs, offset=n_interval.start.offset, datum=cs_datum),
end=hgvs.location.BaseOffsetPosition(base=ce,
offset=n_interval.end.offset,
datum=ce_datum),
uncertain=n_interval.uncertain)
return c_interval
def __str__(self):
if self.ref is None and self.alt is None:
raise HGVSError("RefAlt: ref and alt sequences are both undefined")
# subst and delins
if self.ref is not None and self.alt is not None:
if self.ref == self.alt:
s = "{self.ref}=".format(self=self)
elif len(self.alt) == 1 and len(
self.ref
) == 1 and not self.ref.isdigit(): # don't turn del5insT into 5>T
s = "{self.ref}>{self.alt}".format(self=self)
else:
s = "del{self.ref}ins{self.alt}".format(self=self)
# del case
elif self.ref is not None:
s = "del{self.ref}".format(self=self)
# ins case
else: # self.alt is not None
s = "ins{self.alt}".format(self=self)
return "(" + s + ")" if self.uncertain else s