def _construct_change(var, reverse=False):
"""
Construct mutation description.
@arg var: RawVar object.
@type var: pyparsing.ParseResults
@var reverse: Variant is on the reverse strand.
@type reverse: bool
@return: Description of mutation (without reference and positions).
@rtype: unicode
"""
# Note that the pyparsing parse tree yields `str('')` for nonexisting
# attributes, so we wrap the optional attributes in `unicode()`.
if reverse:
try:
arg1 = unicode(int(var.Arg1))
except ValueError:
arg1 = util.reverse_complement(unicode(var.Arg1))
try:
arg2 = unicode(int(var.Arg2))
except ValueError:
arg2 = util.reverse_complement(unicode(var.Arg2))
else:
arg1 = unicode(var.Arg1)
arg2 = unicode(var.Arg2)
def parse_sequence(seq):
if not seq.Sequence:
raise NotImplementedError('Only explicit sequences are supported '
'for insertions.')
if reverse:
return util.reverse_complement(seq.Sequence)
return seq.Sequence
if var.MutationType == 'subst':
change = '%s>%s' % (arg1, arg2)
elif var.MutationType in ('ins', 'delins'):
if var.SeqList:
if reverse:
seqs = reversed(var.SeqList)
else:
seqs = var.SeqList
insertion = '[' + ';'.join(parse_sequence(seq)
for seq in seqs) + ']'
else:
insertion = parse_sequence(var.Seq)
change = '%s%s' % (var.MutationType, insertion)
else:
change = '%s%s' % (var.MutationType, arg1 or arg2 or '')
return change
def _construct_change(var, reverse=False):
"""
Construct mutation description.
@arg var: RawVar object.
@type var: pyparsing.ParseResults
@var reverse: Variant is on the reverse strand.
@type reverse: bool
@return: Description of mutation (without reference and positions).
@rtype: unicode
"""
# Note that the pyparsing parse tree yields `str('')` for nonexisting
# attributes, so we wrap the optional attributes in `unicode()`.
if reverse:
try:
arg1 = unicode(int(var.Arg1))
except ValueError:
arg1 = util.reverse_complement(unicode(var.Arg1))
try:
arg2 = unicode(int(var.Arg2))
except ValueError:
arg2 = util.reverse_complement(unicode(var.Arg2))
else:
arg1 = unicode(var.Arg1)
arg2 = unicode(var.Arg2)
def parse_sequence(seq):
if not seq.Sequence:
raise NotImplementedError('Only explicit sequences are supported '
'for insertions.')
if reverse:
return util.reverse_complement(seq.Sequence)
return seq.Sequence
if var.MutationType == 'subst':
change = '%s>%s' % (arg1, arg2)
elif var.MutationType in ('ins', 'delins'):
if var.SeqList:
if reverse:
seqs = reversed(var.SeqList)
else:
seqs = var.SeqList
insertion = '[' + ';'.join(parse_sequence(seq)
for seq in seqs) + ']'
else:
insertion = parse_sequence(var.Seq)
change = '%s%s' % (var.MutationType, insertion)
else:
change = '%s%s' % (var.MutationType, arg1 or arg2 or '')
return change
def parse_sequence(seq):
if not seq.Sequence:
raise NotImplementedError('Only explicit sequences are supported '
'for insertions.')
if reverse:
return util.reverse_complement(seq.Sequence)
return seq.Sequence
def parse_sequence(seq):
if not seq.Sequence:
raise NotImplementedError('Only explicit sequences are supported '
'for insertions.')
if reverse:
return util.reverse_complement(seq.Sequence)
return seq.Sequence
def __init__(self, s1, s2, lcp, s1_end, s2_end, DNA=False):
"""
Initialise the class.
@arg s1: A string.
@type s1: unicode
@arg s2: A string.
@type s2: unicode
@arg lcp: The length of the longest common prefix of {s1} and {s2}.
@type lcp: int
@arg s1_end: End of the substring in {s1}.
@type s1_end:
@arg s2_end: End of the substring in {s2}.
@type s2_end: int
@arg DNA:
@type DNA: bool
"""
self.__lcp = lcp
self.__s1 = s1[self.__lcp:s1_end]
self.__s2 = s2[self.__lcp:s2_end]
self.__s2_len = s2_end - lcp
self.__matrix = self.LCSMatrix(self.__s1, self.__s2)
self.__s2_rc = None
self.__matrix_rc = None
if DNA:
self.__s2_rc = reverse_complement(s2[self.__lcp:s2_end])
self.__matrix_rc = self.LCSMatrix(self.__s1, self.__s2_rc)
def inversion(self, pos1, pos2):
"""
Invert a range from non-interbase position pos1 to pos2.
@arg pos1: First nucleotide of the inverted sequence.
@type pos1: int
@arg pos2: Last nucleotide of the inverted sequence.
@type pos2: int
"""
sequence = util.reverse_complement(unicode(self.orig[pos1 - 1:pos2]))
visualisation = ['inversion between %i and %i' % (pos1, pos2)]
visualisation.extend(self._visualise(pos1 - 1, pos2, sequence))
self._output.addOutput('visualisation', visualisation)
self._mutate(pos1 - 1, pos2, sequence)
def inversion(self, pos1, pos2):
"""
Invert a range from non-interbase position pos1 to pos2.
@arg pos1: First nucleotide of the inverted sequence.
@type pos1: int
@arg pos2: Last nucleotide of the inverted sequence.
@type pos2: int
"""
sequence = util.reverse_complement(unicode(self.orig[pos1 - 1:pos2]))
visualisation = ['inversion between %i and %i' % (pos1, pos2)]
visualisation.extend(self._visualise(pos1 - 1, pos2, sequence))
self._output.addOutput('visualisation', visualisation)
self._mutate(pos1 - 1, pos2, sequence)
def __maybeInvert(self, gene, string, string_reverse=None):
"""
Return the reverse-complement of a DNA sequence if the gene is in
the reverse orientation.
@arg gene: Gene
@type gene: object
@arg string: DNA sequence
@type string: unicode
@kwarg string_reverse: DNA sequence to use (if not None) for the
reverse complement.
@return: reverse-complement (if applicable), otherwise return the
original.
@rtype: unicode
"""
if gene.orientation == -1:
if string_reverse:
string = string_reverse
return util.reverse_complement(string)
return string
def __maybeInvert(self, gene, string, string_reverse=None):
"""
Return the reverse-complement of a DNA sequence if the gene is in
the reverse orientation.
@arg gene: Gene
@type gene: object
@arg string: DNA sequence
@type string: unicode
@kwarg string_reverse: DNA sequence to use (if not None) for the
reverse complement.
@return: reverse-complement (if applicable), otherwise return the
original.
@rtype: unicode
"""
if gene.orientation == -1:
if string_reverse:
string = string_reverse
return util.reverse_complement(string)
return string
def name(self, start_g, stop_g, varType, arg1, arg2, roll, arg1_reverse=None, start_fuzzy=False, stop_fuzzy=False):
"""
Generate variant descriptions for all genes, transcripts, etc.
@arg start_g: start position
@type start_g: integer
@arg stop_g: stop position
@type stop_g: integer
@arg varType: variant type
@type varType: unicode
@arg arg1: argument 1 of a raw variant
@type arg1: unicode
@arg arg2: argument 2 of a raw variant
@type arg2: unicode
@arg roll: ???
@type roll: tuple (integer, integer)
@kwarg arg1_reverse: argument 1 to be used on reverse strand
@type arg1_reverse: unicode
@kwarg start_fuzzy: Indicates if start position of variant is fuzzy.
@type start_fuzzy: bool
@kwarg stop_fuzzy: Indicates if stop position of variant is fuzzy.
@type stop_fuzzy: bool
"""
forwardStart = start_g
forwardStop = stop_g
reverseStart = stop_g
reverseStop = start_g
if self.record.orientation == 1:
chromStart = self.record.toChromPos(start_g)
chromStop = self.record.toChromPos(stop_g)
chromArg1 = arg1
chromArg2 = arg2
else:
chromStart = self.record.toChromPos(stop_g)
chromStop = self.record.toChromPos(start_g)
chromArg1 = util.reverse_complement(arg1)
chromArg2 = util.reverse_complement(arg2)
# Todo: Should we use arg1_reverse here?
if roll:
forwardStart += roll[1]
forwardStop += roll[1]
reverseStart -= roll[0]
reverseStop -= roll[0]
if chromStart is not None:
if self.record.orientation == 1:
chromStart += roll[1]
chromStop += roll[1]
else:
chromStart += roll[0]
chromStop += roll[0]
# if
if varType != "subst":
if forwardStart != forwardStop:
# Todo: Fuzzy offsets to genomic positions (see bug #38).
#
# The genomic positioning is problematic. We would like to
# have it in brackets (as fuzzy positions), like the above
# g.(34299_23232)del example.
#
# Now consider a variant c.a-?_b+18del where only the offset
# before the exon is unknown but the offset after the exon is
# exact. Now a genomic description like g.(34299)_23232del
# comes to mind, however, this notation is not allowed by the
# HGVS grammar.
#
# I think all we can do is to treat both positions as fuzzy in
# the genomic description, even if only one of them really is.
#
# Peter thinks the HGVS grammar should at some point be
# updated to allow the brackets around individual locations.
if start_fuzzy or stop_fuzzy:
self.record.addToDescription("(%s_%s)%s%s" % (forwardStart, forwardStop, varType, arg1))
self.record.addToChromDescription("(%s_%s)%s%s" % (chromStart, chromStop, varType, chromArg1))
else:
self.record.addToDescription("%s_%s%s%s" % (forwardStart, forwardStop, varType, arg1))
self.record.addToChromDescription("%s_%s%s%s" % (chromStart, chromStop, varType, chromArg1))
# if
else:
if start_fuzzy or stop_fuzzy:
# Todo: Current HGVS does not allow for () around single
# positions, only around ranges (see above and #38).
self.record.addToDescription("(%s)%s%s" % (forwardStart, varType, arg1))
self.record.addToChromDescription("(%s)%s%s" % (chromStart, varType, chromArg1))
else:
self.record.addToDescription("%s%s%s" % (forwardStart, varType, arg1))
self.record.addToChromDescription("%s%s%s" % (chromStart, varType, chromArg1))
# else
# if
else:
if start_fuzzy or stop_fuzzy:
# Todo: Current HGVS does not allow for () around single
# positions, only around ranges (see above and #38).
self.record.addToDescription("(%s)%c>%c" % (forwardStart, arg1, arg2))
self.record.addToChromDescription("(%s)%c>%c" % (chromStart, chromArg1, chromArg2))
else:
self.record.addToDescription("%s%c>%c" % (forwardStart, arg1, arg2))
self.record.addToChromDescription("%s%c>%c" % (chromStart, chromArg1, chromArg2))
for i in self.record.geneList:
for j in i.transcriptList:
if j.CM:
orientedStart = forwardStart
orientedStop = forwardStop
if i.orientation == -1:
orientedStart = reverseStart
orientedStop = reverseStop
# if
# Turn of translation to protein if we hit splice sites.
# For the current transcript, this is handled with more
# care in variantchecker.py.
if not j.current and util.over_splice_site(orientedStart, orientedStop, j.CM.RNA):
j.translate = False
# And check whether the variant hits CDS start.
if j.molType == "c" and forwardStop >= j.CM.x2g(1, 0) and forwardStart <= j.CM.x2g(3, 0):
self.__output.addMessage(
__file__,
2,
"WSTART",
"Mutation in start codon of gene %s transcript " "%s." % (i.name, j.name),
)
if not j.current:
j.translate = False
# FIXME Check whether the variant hits a splice site.
if varType != "subst":
if orientedStart != orientedStop:
if (start_fuzzy or stop_fuzzy) and not j.current:
# Don't generate descriptions on transcripts
# other than the current in the case of fuzzy
# positions.
j.cancelDescription()
else:
j.addToDescription(
"%s_%s%s%s"
% (
j.CM.g2c(orientedStart, start_fuzzy),
j.CM.g2c(orientedStop, stop_fuzzy),
varType,
self.__maybeInvert(i, arg1, arg1_reverse),
)
)
self.checkIntron(i, j, orientedStart)
self.checkIntron(i, j, orientedStop)
# if
else:
if start_fuzzy and not j.current:
# Don't generate descriptions on transcripts
# other than the current in the case of fuzzy
# positions.
j.cancelDescription()
else:
j.addToDescription(
"%s%s%s"
% (
j.CM.g2c(orientedStart, start_fuzzy),
varType,
self.__maybeInvert(i, arg1, arg1_reverse),
)
)
self.checkIntron(i, j, orientedStart)
# else
# if
else:
if start_fuzzy and not j.current:
# Don't generate descriptions on transcripts
# other than the current in the case of fuzzy
# positions.
j.cancelDescription()
else:
j.addToDescription(
"%s%c>%c"
% (
j.CM.g2c(orientedStart, start_fuzzy),
self.__maybeInvert(i, arg1, arg1_reverse),
self.__maybeInvert(i, arg2),
)
)
self.checkIntron(i, j, orientedStart)
def name(self,
start_g,
stop_g,
varType,
arg1,
arg2,
roll,
arg1_reverse=None,
start_fuzzy=False,
stop_fuzzy=False):
"""
Generate variant descriptions for all genes, transcripts, etc.
@arg start_g: start position
@type start_g: integer
@arg stop_g: stop position
@type stop_g: integer
@arg varType: variant type
@type varType: unicode
@arg arg1: argument 1 of a raw variant
@type arg1: unicode
@arg arg2: argument 2 of a raw variant
@type arg2: unicode
@arg roll: ???
@type roll: tuple (integer, integer)
@kwarg arg1_reverse: argument 1 to be used on reverse strand
@type arg1_reverse: unicode
@kwarg start_fuzzy: Indicates if start position of variant is fuzzy.
@type start_fuzzy: bool
@kwarg stop_fuzzy: Indicates if stop position of variant is fuzzy.
@type stop_fuzzy: bool
"""
forwardStart = start_g
forwardStop = stop_g
reverseStart = stop_g
reverseStop = start_g
if self.record.orientation == 1:
chromStart = self.record.toChromPos(start_g)
chromStop = self.record.toChromPos(stop_g)
chromArg1 = arg1
chromArg2 = arg2
else:
chromStart = self.record.toChromPos(stop_g)
chromStop = self.record.toChromPos(start_g)
chromArg1 = util.reverse_complement(arg1)
chromArg2 = util.reverse_complement(arg2)
# Todo: Should we use arg1_reverse here?
if roll:
forwardStart += roll[1]
forwardStop += roll[1]
reverseStart -= roll[0]
reverseStop -= roll[0]
if chromStart is not None:
if self.record.orientation == 1:
chromStart += roll[1]
chromStop += roll[1]
else:
chromStart += roll[0]
chromStop += roll[0]
#if
if varType != "subst":
if forwardStart != forwardStop:
# Todo: Fuzzy offsets to genomic positions (see bug #38).
#
# The genomic positioning is problematic. We would like to
# have it in brackets (as fuzzy positions), like the above
# g.(34299_23232)del example.
#
# Now consider a variant c.a-?_b+18del where only the offset
# before the exon is unknown but the offset after the exon is
# exact. Now a genomic description like g.(34299)_23232del
# comes to mind, however, this notation is not allowed by the
# HGVS grammar.
#
# I think all we can do is to treat both positions as fuzzy in
# the genomic description, even if only one of them really is.
#
# Peter thinks the HGVS grammar should at some point be
# updated to allow the brackets around individual locations.
if start_fuzzy or stop_fuzzy:
self.record.addToDescription(
"(%s_%s)%s%s" %
(forwardStart, forwardStop, varType, arg1))
self.record.addToChromDescription(
"(%s_%s)%s%s" %
(chromStart, chromStop, varType, chromArg1))
else:
self.record.addToDescription(
"%s_%s%s%s" %
(forwardStart, forwardStop, varType, arg1))
self.record.addToChromDescription(
"%s_%s%s%s" %
(chromStart, chromStop, varType, chromArg1))
#if
else:
if start_fuzzy or stop_fuzzy:
# Todo: Current HGVS does not allow for () around single
# positions, only around ranges (see above and #38).
self.record.addToDescription("(%s)%s%s" %
(forwardStart, varType, arg1))
self.record.addToChromDescription(
"(%s)%s%s" % (chromStart, varType, chromArg1))
else:
self.record.addToDescription("%s%s%s" %
(forwardStart, varType, arg1))
self.record.addToChromDescription(
"%s%s%s" % (chromStart, varType, chromArg1))
#else
#if
else:
if start_fuzzy or stop_fuzzy:
# Todo: Current HGVS does not allow for () around single
# positions, only around ranges (see above and #38).
self.record.addToDescription("(%s)%c>%c" %
(forwardStart, arg1, arg2))
self.record.addToChromDescription(
"(%s)%c>%c" % (chromStart, chromArg1, chromArg2))
else:
self.record.addToDescription("%s%c>%c" %
(forwardStart, arg1, arg2))
self.record.addToChromDescription(
"%s%c>%c" % (chromStart, chromArg1, chromArg2))
for i in self.record.geneList:
for j in i.transcriptList:
if j.CM:
orientedStart = forwardStart
orientedStop = forwardStop
if i.orientation == -1:
orientedStart = reverseStart
orientedStop = reverseStop
#if
# Turn of translation to protein if we hit splice sites.
# For the current transcript, this is handled with more
# care in variantchecker.py.
if not j.current and \
util.over_splice_site(orientedStart, orientedStop,
j.CM.RNA):
j.translate = False
# And check whether the variant hits CDS start.
if j.molType == 'c' and forwardStop >= j.CM.x2g(1, 0) \
and forwardStart <= j.CM.x2g(3, 0) :
self.__output.addMessage(__file__, 2, "WSTART",
"Mutation in start codon of gene %s transcript " \
"%s." % (i.name, j.name))
if not j.current:
j.translate = False
# FIXME Check whether the variant hits a splice site.
if varType != "subst":
if orientedStart != orientedStop:
if (start_fuzzy or stop_fuzzy) and not j.current:
# Don't generate descriptions on transcripts
# other than the current in the case of fuzzy
# positions.
j.cancelDescription()
else:
j.addToDescription(
"%s_%s%s%s" %
(j.CM.g2c(orientedStart, start_fuzzy),
j.CM.g2c(orientedStop,
stop_fuzzy), varType,
self.__maybeInvert(i, arg1,
arg1_reverse)))
self.checkIntron(i, j, orientedStart)
self.checkIntron(i, j, orientedStop)
#if
else:
if start_fuzzy and not j.current:
# Don't generate descriptions on transcripts
# other than the current in the case of fuzzy
# positions.
j.cancelDescription()
else:
j.addToDescription(
"%s%s%s" % (j.CM.g2c(
orientedStart, start_fuzzy), varType,
self.__maybeInvert(
i, arg1, arg1_reverse)))
self.checkIntron(i, j, orientedStart)
#else
#if
else:
if start_fuzzy and not j.current:
# Don't generate descriptions on transcripts
# other than the current in the case of fuzzy
# positions.
j.cancelDescription()
else:
j.addToDescription(
"%s%c>%c" %
(j.CM.g2c(orientedStart, start_fuzzy),
self.__maybeInvert(i, arg1, arg1_reverse),
self.__maybeInvert(i, arg2)))
self.checkIntron(i, j, orientedStart)
