def common_ancestor(self, targets, *more_targets):
"""Most recent common ancestor (clade) of all the given targets.
Edge cases:
- If no target is given, returns self.root
- If 1 target is given, returns the target
- If any target is not found in this tree, raises a ValueError
"""
paths = [self.get_path(t)
for t in _combine_args(targets, *more_targets)]
# Validation -- otherwise izip throws a spooky error below
for p, t in zip(paths, targets):
if p is None:
raise ValueError("target %s is not in this tree" % repr(t))
mrca = self.root
for level in zip(*paths):
ref = level[0]
for other in level[1:]:
if ref is not other:
break
else:
mrca = ref
if ref is not mrca:
break
return mrca
def common_ancestor(self, targets, *more_targets):
"""Most recent common ancestor (clade) of all the given targets.
Edge cases:
- If no target is given, returns self.root
- If 1 target is given, returns the target
- If any target is not found in this tree, raises a ValueError
"""
paths = [
self.get_path(t) for t in _combine_args(targets, *more_targets)
]
# Validation -- otherwise izip throws a spooky error below
for p, t in zip(paths, targets):
if p is None:
raise ValueError("target %s is not in this tree" % repr(t))
mrca = self.root
for level in zip(*paths):
ref = level[0]
for other in level[1:]:
if ref is not other:
break
else:
mrca = ref
if ref is not mrca:
break
return mrca
def _get_inter_coords(coords, strand=1):
"""From the given pairs of coordinates, returns a list of pairs
covering the intervening ranges."""
# adapted from Python's itertools guide
# if strand is -1, adjust coords to the ends and starts are chained
if strand == -1:
sorted_coords = [(max(a, b), min(a, b)) for a, b in coords]
inter_coords = list(chain(*sorted_coords))[1:-1]
return list(zip(inter_coords[1::2], inter_coords[::2]))
else:
inter_coords = list(chain(*coords))[1:-1]
return list(zip(inter_coords[::2], inter_coords[1::2]))
def _get_inter_coords(coords, strand=1):
"""From the given pairs of coordinates, returns a list of pairs
covering the intervening ranges."""
# adapted from Python's itertools guide
# if strand is -1, adjust coords to the ends and starts are chained
if strand == -1:
sorted_coords = [(max(a, b), min(a, b)) for a, b in coords]
inter_coords = list(chain(*sorted_coords))[1:-1]
return list(zip(inter_coords[1::2], inter_coords[::2]))
else:
inter_coords = list(chain(*coords))[1:-1]
return list(zip(inter_coords[::2], inter_coords[1::2]))
def _reorient_starts(starts, blksizes, seqlen, strand):
"""Reorients block starts into the opposite strand's coordinates.
Arguments:
starts -- List of integers, start coordinates.
start -- Integer, 'Q start' or 'T start' column
blksizes -- List of integers, block sizes.
seqlen -- Integer of total sequence length.
strand -- Integer denoting sequence strand.
"""
assert len(starts) == len(blksizes), \
"Unequal start coordinates and block sizes list (%r vs %r)" \
% (len(starts), len(blksizes))
# see: http://genome.ucsc.edu/goldenPath/help/blatSpec.html
# no need to reorient if it's already the positive strand
if strand >= 0:
return starts
else:
# the plus-oriented coordinate is calculated by this:
# plus_coord = length - minus_coord - block_size
return [
seqlen - start - blksize
for start, blksize in zip(starts, blksizes)
]
def _comp_intron_lens(seq_type, inter_blocks, raw_inter_lens):
"""Returns the length of introns between fragments."""
# set opposite type, for setting introns
opp_type = 'hit' if seq_type == 'query' else 'query'
# list of flags to denote if an intron follows a block
# it reads e.g. this line:
# "ATGTT{TT} >>>> Target Intron 1 >>>> {G}TGTGTGTACATT"
# and sets the opposing sequence type's intron (since this
# line is present on the opposite sequence type line)
has_intron_after = ['Intron' in x[seq_type] for x in inter_blocks]
assert len(has_intron_after) == len(raw_inter_lens)
# create list containing coord adjustments incorporating
# intron lengths
inter_lens = []
for flag, parsed_len in zip(has_intron_after, raw_inter_lens):
if flag:
# joint introns
if all(parsed_len[:2]):
# intron len is [0] if opp_type is query, otherwise it's [1]
intron_len = int(parsed_len[0]) if opp_type == 'query' \
else int(parsed_len[1])
# single hit/query introns
elif parsed_len[2]:
intron_len = int(parsed_len[2])
else:
raise ValueError("Unexpected intron parsing "
"result: %r" % parsed_len)
else:
intron_len = 0
inter_lens.append(intron_len)
return inter_lens
def randomized(cls, taxa, branch_length=1.0, branch_stdev=None):
"""Create a randomized bifurcating tree given a list of taxa.
:param taxa: Either an integer specifying the number of taxa to create
(automatically named taxon#), or an iterable of taxon names, as
strings.
:returns: a tree of the same type as this class.
"""
if isinstance(taxa, int):
taxa = ['taxon%s' % (i+1) for i in range(taxa)]
elif hasattr(taxa, '__iter__'):
taxa = list(taxa)
else:
raise TypeError("taxa argument must be integer (# taxa) or "
"iterable of taxon names.")
rtree = cls()
terminals = [rtree.root]
while len(terminals) < len(taxa):
newsplit = random.choice(terminals)
newsplit.split(branch_length=branch_length)
newterms = newsplit.clades
if branch_stdev:
# Add some noise to the branch lengths
for nt in newterms:
nt.branch_length = max(0,
random.gauss(branch_length, branch_stdev))
terminals.remove(newsplit)
terminals.extend(newterms)
# Distribute taxon labels randomly
random.shuffle(taxa)
for node, name in zip(terminals, taxa):
node.name = name
return rtree
def randomized(cls, taxa, branch_length=1.0, branch_stdev=None):
"""Create a randomized bifurcating tree given a list of taxa.
:param taxa: Either an integer specifying the number of taxa to create
(automatically named taxon#), or an iterable of taxon names, as
strings.
:returns: a tree of the same type as this class.
"""
if isinstance(taxa, int):
taxa = ['taxon%s' % (i + 1) for i in range(taxa)]
elif hasattr(taxa, '__iter__'):
taxa = list(taxa)
else:
raise TypeError("taxa argument must be integer (# taxa) or "
"iterable of taxon names.")
rtree = cls()
terminals = [rtree.root]
while len(terminals) < len(taxa):
newsplit = random.choice(terminals)
newsplit.split(branch_length=branch_length)
newterms = newsplit.clades
if branch_stdev:
# Add some noise to the branch lengths
for nt in newterms:
nt.branch_length = max(
0, random.gauss(branch_length, branch_stdev))
terminals.remove(newsplit)
terminals.extend(newterms)
# Distribute taxon labels randomly
random.shuffle(taxa)
for node, name in zip(terminals, taxa):
node.name = name
return rtree
def _comp_intron_lens(seq_type, inter_blocks, raw_inter_lens):
"""Returns the length of introns between fragments."""
# set opposite type, for setting introns
opp_type = 'hit' if seq_type == 'query' else 'query'
# list of flags to denote if an intron follows a block
# it reads e.g. this line:
# "ATGTT{TT} >>>> Target Intron 1 >>>> {G}TGTGTGTACATT"
# and sets the opposing sequence type's intron (since this
# line is present on the opposite sequence type line)
has_intron_after = ['Intron' in x[seq_type] for x in
inter_blocks]
assert len(has_intron_after) == len(raw_inter_lens)
# create list containing coord adjustments incorporating
# intron lengths
inter_lens = []
for flag, parsed_len in zip(has_intron_after, raw_inter_lens):
if flag:
# joint introns
if all(parsed_len[:2]):
# intron len is [0] if opp_type is query, otherwise it's [1]
intron_len = int(parsed_len[0]) if opp_type == 'query' \
else int(parsed_len[1])
# single hit/query introns
elif parsed_len[2]:
intron_len = int(parsed_len[2])
else:
raise ValueError("Unexpected intron parsing "
"result: %r" % parsed_len)
else:
intron_len = 0
inter_lens.append(intron_len)
return inter_lens
def __init__(self, line=None):
self.chromat_file = ''
self.phd_file = ''
self.time = ''
self.chem = ''
self.dye = ''
self.template = ''
self.direction = ''
if line:
tags = ['CHROMAT_FILE', 'PHD_FILE', 'TIME', 'CHEM', 'DYE', 'TEMPLATE', 'DIRECTION']
poss = [line.find(x) for x in tags]
tagpos = dict(zip(poss, tags))
if -1 in tagpos:
del tagpos[-1]
ps = sorted(tagpos) # the keys
for (p1, p2) in zip(ps, ps[1:]+[len(line)+1]):
setattr(self, tagpos[p1].lower(), line[p1+len(tagpos[p1])+1:p2].strip())
def _flip_codons(codon_seq, target_seq):
"""Flips the codon characters from one seq to another."""
a, b = '', ''
for char1, char2 in zip(codon_seq, target_seq):
# no need to do anything if the codon seq line has nothing
if char1 == ' ':
a += char1
b += char2
else:
a += char2
b += char1
return a, b
def _gaf10iterator(handle):
for inline in handle:
if inline[0] == '!':
continue
inrec = inline.rstrip('\n').split('\t')
if len(inrec) == 1:
continue
inrec[3] = inrec[3].split('|') #Qualifier
inrec[5] = inrec[5].split('|') # DB:reference(s)
inrec[7] = inrec[7].split('|') # With || From
inrec[10] = inrec[10].split('|') # Synonym
inrec[12] = inrec[12].split('|') # Taxon
yield dict(zip(GAF10FIELDS, inrec))
def _flip_codons(codon_seq, target_seq):
"""Flips the codon characters from one seq to another."""
a, b = '', ''
for char1, char2 in zip(codon_seq, target_seq):
# no need to do anything if the codon seq line has nothing
if char1 == ' ':
a += char1
b += char2
else:
a += char2
b += char1
return a, b
def _gaf10iterator(handle):
for inline in handle:
if inline[0] == '!':
continue
inrec = inline.rstrip('\n').split('\t')
if len(inrec) == 1:
continue
inrec[3] = inrec[3].split('|') #Qualifier
inrec[5] = inrec[5].split('|') # DB:reference(s)
inrec[7] = inrec[7].split('|') # With || From
inrec[10] = inrec[10].split('|') # Synonym
inrec[12] = inrec[12].split('|') # Taxon
yield dict(zip(GAF10FIELDS, inrec))
def _gpi10iterator(handle):
"""Read GPI 1.0 format files (PRIVATE).
This iterator is used to read a gp_information.goa_uniprot
file which is in the GPI 1.0 format.
"""
for inline in handle:
if inline[0] == '!':
continue
inrec = inline.rstrip('\n').split('\t')
if len(inrec) == 1:
continue
inrec[5] = inrec[5].split('|') # DB_Object_Synonym(s)
inrec[8] = inrec[8].split('|') # Annotation_Target_Set
yield dict(zip(GPI10FIELDS, inrec))
def _gpi10iterator(handle):
"""Read GPI 1.0 format files (PRIVATE).
This iterator is used to read a gp_information.goa_uniprot
file which is in the GPI 1.0 format.
"""
for inline in handle:
if inline[0] == '!':
continue
inrec = inline.rstrip('\n').split('\t')
if len(inrec) == 1:
continue
inrec[5] = inrec[5].split('|') # DB_Object_Synonym(s)
inrec[8] = inrec[8].split('|') # Annotation_Target_Set
yield dict(zip(GPI10FIELDS, inrec))
def _get_coords(filename):
alb = file(filename)
start_line = None
end_line = None
for line in alb:
if line.startswith("["):
if not start_line:
start_line = line # rstrip not needed
else:
end_line = line
if end_line is None: # sequence is too short
return [(0, 0), (0, 0)]
return list(zip(*map(_alb_line2coords, [start_line, end_line]))) # returns [(start0, end0), (start1, end1)]
def _gpi11iterator(handle):
"""Read GPI 1.0 format files (PRIVATE).
This iterator is used to read a gp_information.goa_uniprot
file which is in the GPI 1.0 format.
"""
for inline in handle:
if inline[0] == '!':
continue
inrec = inline.rstrip('\n').split('\t')
if len(inrec) == 1:
continue
inrec[2] = inrec[2].split('|') # DB_Object_Name
inrec[3] = inrec[3].split('|') # DB_Object_Synonym(s)
inrec[7] = inrec[7].split('|') # DB_Xref(s)
inrec[8] = inrec[8].split('|') # Properties
yield dict(zip(GPI11FIELDS, inrec))
def _gpi11iterator(handle):
"""Read GPI 1.0 format files (PRIVATE).
This iterator is used to read a gp_information.goa_uniprot
file which is in the GPI 1.0 format.
"""
for inline in handle:
if inline[0] == '!':
continue
inrec = inline.rstrip('\n').split('\t')
if len(inrec) == 1:
continue
inrec[2] = inrec[2].split('|') # DB_Object_Name
inrec[3] = inrec[3].split('|') # DB_Object_Synonym(s)
inrec[7] = inrec[7].split('|') # DB_Xref(s)
inrec[8] = inrec[8].split('|') # Properties
yield dict(zip(GPI11FIELDS, inrec))
def _get_coords(filename):
alb = file(filename)
start_line = None
end_line = None
for line in alb:
if line.startswith("["):
if not start_line:
start_line = line # rstrip not needed
else:
end_line = line
if end_line is None: # sequence is too short
return [(0, 0), (0, 0)]
return list(zip(*map(_alb_line2coords, [start_line, end_line]))
) # returns [(start0, end0), (start1, end1)]
def _gpa11iterator(handle):
"""Read GPA 1.1 format files (PRIVATE).
This iterator is used to read a gp_association.goa_uniprot
file which is in the GPA 1.1 format. Do not call directly. Rather
use the gpa_iterator function
"""
for inline in handle:
if inline[0] == '!':
continue
inrec = inline.rstrip('\n').split('\t')
if len(inrec) == 1:
continue
inrec[2] = inrec[2].split('|') # Qualifier
inrec[4] = inrec[4].split('|') # DB:Reference(s)
inrec[6] = inrec[6].split('|') # With
inrec[10] = inrec[10].split('|') # Annotation extension
yield dict(zip(GPA11FIELDS, inrec))
def _gpa11iterator(handle):
"""Read GPA 1.1 format files (PRIVATE).
This iterator is used to read a gp_association.goa_uniprot
file which is in the GPA 1.1 format. Do not call directly. Rather
use the gpa_iterator function
"""
for inline in handle:
if inline[0] == '!':
continue
inrec = inline.rstrip('\n').split('\t')
if len(inrec) == 1:
continue
inrec[2] = inrec[2].split('|') # Qualifier
inrec[4] = inrec[4].split('|') # DB:Reference(s)
inrec[6] = inrec[6].split('|') # With
inrec[10] = inrec[10].split('|') # Annotation extension
yield dict(zip(GPA11FIELDS, inrec))
def _gaf20byproteiniterator(handle):
cur_id = None
id_rec_list = []
for inline in handle:
if inline[0] == '!':
continue
inrec = inline.rstrip('\n').split('\t')
if len(inrec) == 1:
continue
inrec[3] = inrec[3].split('|') #Qualifier
inrec[5] = inrec[5].split('|') # DB:reference(s)
inrec[7] = inrec[7].split('|') # With || From
inrec[10] = inrec[10].split('|') # Synonym
inrec[12] = inrec[12].split('|') # Taxon
cur_rec = dict(zip(GAF20FIELDS, inrec))
if cur_rec['DB_Object_ID'] != cur_id and cur_id:
ret_list = copy.copy(id_rec_list)
id_rec_list = [cur_rec]
cur_id = cur_rec['DB_Object_ID']
yield ret_list
else:
cur_id = cur_rec['DB_Object_ID']
id_rec_list.append(cur_rec)
def _reorient_starts(starts, blksizes, seqlen, strand):
"""Reorients block starts into the opposite strand's coordinates.
Arguments:
starts -- List of integers, start coordinates.
start -- Integer, 'Q start' or 'T start' column
blksizes -- List of integers, block sizes.
seqlen -- Integer of total sequence length.
strand -- Integer denoting sequence strand.
"""
assert len(starts) == len(blksizes), \
"Unequal start coordinates and block sizes list (%r vs %r)" \
% (len(starts), len(blksizes))
# see: http://genome.ucsc.edu/goldenPath/help/blatSpec.html
# no need to reorient if it's already the positive strand
if strand >= 0:
return starts
else:
# the plus-oriented coordinate is calculated by this:
# plus_coord = length - minus_coord - block_size
return [seqlen - start - blksize for
start, blksize in zip(starts, blksizes)]
def _gaf20byproteiniterator(handle):
cur_id = None
id_rec_list = []
for inline in handle:
if inline[0] == '!':
continue
inrec = inline.rstrip('\n').split('\t')
if len(inrec) == 1:
continue
inrec[3] = inrec[3].split('|') #Qualifier
inrec[5] = inrec[5].split('|') # DB:reference(s)
inrec[7] = inrec[7].split('|') # With || From
inrec[10] = inrec[10].split('|') # Synonym
inrec[12] = inrec[12].split('|') # Taxon
cur_rec = dict(zip(GAF20FIELDS, inrec))
if cur_rec['DB_Object_ID'] != cur_id and cur_id:
ret_list = copy.copy(id_rec_list)
id_rec_list = [cur_rec]
cur_id = cur_rec['DB_Object_ID']
yield ret_list
else:
cur_id = cur_rec['DB_Object_ID']
id_rec_list.append(cur_rec)
def AbiIterator(handle, alphabet=None, trim=False):
"""Iterator for the Abi file format.
"""
# raise exception is alphabet is not dna
if alphabet is not None:
if isinstance(Alphabet._get_base_alphabet(alphabet),
Alphabet.ProteinAlphabet):
raise ValueError(
"Invalid alphabet, ABI files do not hold proteins.")
if isinstance(Alphabet._get_base_alphabet(alphabet),
Alphabet.RNAAlphabet):
raise ValueError("Invalid alphabet, ABI files do not hold RNA.")
# raise exception if handle mode is not 'rb'
if hasattr(handle, 'mode'):
if set('rb') != set(handle.mode.lower()):
raise ValueError("ABI files has to be opened in 'rb' mode.")
# check if input file is a valid Abi file
handle.seek(0)
marker = handle.read(4)
if not marker:
# handle empty file gracefully
raise StopIteration
if marker != _as_bytes('ABIF'):
raise IOError('File should start ABIF, not %r' % marker)
# dirty hack for handling time information
times = {'RUND1': '', 'RUND2': '', 'RUNT1': '', 'RUNT2': '', }
# initialize annotations
annot = dict(zip(_EXTRACT.values(), [None] * len(_EXTRACT)))
# parse header and extract data from directories
header = struct.unpack(_HEADFMT,
handle.read(struct.calcsize(_HEADFMT)))
for tag_name, tag_number, tag_data in _abi_parse_header(header, handle):
# stop iteration if all desired tags have been extracted
# 4 tags from _EXTRACT + 2 time tags from _SPCTAGS - 3,
# and seq, qual, id
# todo
key = tag_name + str(tag_number)
# PBAS2 is base-called sequence
if key == 'PBAS2':
seq = tag_data
ambigs = 'KYWMRS'
if alphabet is None:
if set(seq).intersection(ambigs):
alphabet = ambiguous_dna
else:
alphabet = unambiguous_dna
# PCON2 is quality values of base-called sequence
elif key == 'PCON2':
qual = [ord(val) for val in tag_data]
# SMPL1 is sample id entered before sequencing run
elif key == 'SMPL1':
sample_id = tag_data
elif key in times:
times[key] = tag_data
else:
# extract sequence annotation as defined in _EXTRACT
if key in _EXTRACT:
annot[_EXTRACT[key]] = tag_data
# set time annotations
annot['run_start'] = '%s %s' % (times['RUND1'], times['RUNT1'])
annot['run_finish'] = '%s %s' % (times['RUND2'], times['RUNT2'])
# use the file name as SeqRecord.name if available
try:
file_name = basename(handle.name).replace('.ab1', '')
except:
file_name = ""
record = SeqRecord(Seq(seq, alphabet),
id=sample_id, name=file_name,
description='',
annotations=annot,
letter_annotations={'phred_quality': qual})
if not trim:
yield record
else:
yield _abi_trim(record)
def AbiIterator(handle, alphabet=None, trim=False):
"""Iterator for the Abi file format.
"""
# raise exception is alphabet is not dna
if alphabet is not None:
if isinstance(Alphabet._get_base_alphabet(alphabet), Alphabet.ProteinAlphabet):
raise ValueError("Invalid alphabet, ABI files do not hold proteins.")
if isinstance(Alphabet._get_base_alphabet(alphabet), Alphabet.RNAAlphabet):
raise ValueError("Invalid alphabet, ABI files do not hold RNA.")
# raise exception if handle mode is not 'rb'
if hasattr(handle, "mode"):
if set("rb") != set(handle.mode.lower()):
raise ValueError("ABI files has to be opened in 'rb' mode.")
# check if input file is a valid Abi file
handle.seek(0)
marker = handle.read(4)
if not marker:
# handle empty file gracefully
raise StopIteration
if marker != _as_bytes("ABIF"):
raise IOError("File should start ABIF, not %r" % marker)
# dirty hack for handling time information
times = {"RUND1": "", "RUND2": "", "RUNT1": "", "RUNT2": ""}
# initialize annotations
annot = dict(zip(_EXTRACT.values(), [None] * len(_EXTRACT)))
# parse header and extract data from directories
header = struct.unpack(_HEADFMT, handle.read(struct.calcsize(_HEADFMT)))
for tag_name, tag_number, tag_data in _abi_parse_header(header, handle):
# stop iteration if all desired tags have been extracted
# 4 tags from _EXTRACT + 2 time tags from _SPCTAGS - 3,
# and seq, qual, id
# todo
key = tag_name + str(tag_number)
# PBAS2 is base-called sequence
if key == "PBAS2":
seq = tag_data
ambigs = "KYWMRS"
if alphabet is None:
if set(seq).intersection(ambigs):
alphabet = ambiguous_dna
else:
alphabet = unambiguous_dna
# PCON2 is quality values of base-called sequence
elif key == "PCON2":
qual = [ord(val) for val in tag_data]
# SMPL1 is sample id entered before sequencing run
elif key == "SMPL1":
sample_id = tag_data
elif key in times:
times[key] = tag_data
else:
# extract sequence annotation as defined in _EXTRACT
if key in _EXTRACT:
annot[_EXTRACT[key]] = tag_data
# set time annotations
annot["run_start"] = "%s %s" % (times["RUND1"], times["RUNT1"])
annot["run_finish"] = "%s %s" % (times["RUND2"], times["RUNT2"])
# use the file name as SeqRecord.name if available
try:
file_name = basename(handle.name).replace(".ab1", "")
except:
file_name = ""
record = SeqRecord(
Seq(seq, alphabet),
id=sample_id,
name=file_name,
description="",
annotations=annot,
letter_annotations={"phred_quality": qual},
)
if not trim:
yield record
else:
yield _abi_trim(record)
def _create_hsp(hid, qid, psl):
# protein flag
is_protein = _is_protein(psl)
# strand
#if query is protein, strand is 0
if is_protein:
qstrand = 0
else:
qstrand = 1 if psl['strand'][0] == '+' else -1
# try to get hit strand, if it exists
try:
hstrand = 1 if psl['strand'][1] == '+' else -1
except IndexError:
hstrand = 1 # hit strand defaults to plus
# query block starts
qstarts = _reorient_starts(psl['qstarts'],
psl['blocksizes'], psl['qsize'], qstrand)
# hit block starts
if len(psl['strand']) == 2:
hstarts = _reorient_starts(psl['tstarts'],
psl['blocksizes'], psl['tsize'], hstrand)
else:
hstarts = psl['tstarts']
# set query and hit coords
# this assumes each block has no gaps (which seems to be the case)
assert len(qstarts) == len(hstarts) == len(psl['blocksizes'])
query_range_all = list(zip(qstarts, [x + y for x, y in
zip(qstarts, psl['blocksizes'])]))
hit_range_all = list(zip(hstarts, [x + y for x, y in
zip(hstarts, psl['blocksizes'])]))
# check length of sequences and coordinates, all must match
if 'tseqs' in psl and 'qseqs' in psl:
assert len(psl['tseqs']) == len(psl['qseqs']) == \
len(query_range_all) == len(hit_range_all)
else:
assert len(query_range_all) == len(hit_range_all)
frags = []
# iterating over query_range_all, but hit_range_all works just as well
for idx, qcoords in enumerate(query_range_all):
hseqlist = psl.get('tseqs')
hseq = '' if not hseqlist else hseqlist[idx]
qseqlist = psl.get('qseqs')
qseq = '' if not qseqlist else qseqlist[idx]
frag = HSPFragment(hid, qid, hit=hseq, query=qseq)
# set alphabet
frag.alphabet = generic_dna
# set coordinates
frag.query_start = qcoords[0]
frag.query_end = qcoords[1]
frag.hit_start = hit_range_all[idx][0]
frag.hit_end = hit_range_all[idx][1]
# and strands
frag.query_strand = qstrand
frag.hit_strand = hstrand
frags.append(frag)
# create hsp object
hsp = HSP(frags)
# check if start and end are set correctly
assert hsp.query_start == psl['qstart']
assert hsp.query_end == psl['qend']
assert hsp.hit_start == psl['tstart']
assert hsp.hit_end == psl['tend']
# and check block spans as well
assert hsp.query_span_all == hsp.hit_span_all == psl['blocksizes']
# set its attributes
hsp.match_num = psl['matches']
hsp.mismatch_num = psl['mismatches']
hsp.match_rep_num = psl['repmatches']
hsp.n_num = psl['ncount']
hsp.query_gapopen_num = psl['qnuminsert']
hsp.query_gap_num = psl['qbaseinsert']
hsp.hit_gapopen_num = psl['tnuminsert']
hsp.hit_gap_num = psl['tbaseinsert']
hsp.ident_num = psl['matches'] + psl['repmatches']
hsp.gapopen_num = psl['qnuminsert'] + psl['tnuminsert']
hsp.gap_num = psl['qbaseinsert'] + psl['tbaseinsert']
hsp.query_is_protein = is_protein
hsp.ident_pct = 100.0 - _calc_millibad(psl, is_protein) * 0.1
hsp.score = _calc_score(psl, is_protein)
# helper flag, for writing
hsp._has_hit_strand = len(psl['strand']) == 2
return hsp
def _create_hsp(hid, qid, psl):
# protein flag
is_protein = _is_protein(psl)
# strand
#if query is protein, strand is 0
if is_protein:
qstrand = 0
else:
qstrand = 1 if psl['strand'][0] == '+' else -1
# try to get hit strand, if it exists
try:
hstrand = 1 if psl['strand'][1] == '+' else -1
except IndexError:
hstrand = 1 # hit strand defaults to plus
# query block starts
qstarts = _reorient_starts(psl['qstarts'], psl['blocksizes'], psl['qsize'],
qstrand)
# hit block starts
if len(psl['strand']) == 2:
hstarts = _reorient_starts(psl['tstarts'], psl['blocksizes'],
psl['tsize'], hstrand)
else:
hstarts = psl['tstarts']
# set query and hit coords
# this assumes each block has no gaps (which seems to be the case)
assert len(qstarts) == len(hstarts) == len(psl['blocksizes'])
query_range_all = list(
zip(qstarts, [x + y for x, y in zip(qstarts, psl['blocksizes'])]))
hit_range_all = list(
zip(hstarts, [x + y for x, y in zip(hstarts, psl['blocksizes'])]))
# check length of sequences and coordinates, all must match
if 'tseqs' in psl and 'qseqs' in psl:
assert len(psl['tseqs']) == len(psl['qseqs']) == \
len(query_range_all) == len(hit_range_all)
else:
assert len(query_range_all) == len(hit_range_all)
frags = []
# iterating over query_range_all, but hit_range_all works just as well
for idx, qcoords in enumerate(query_range_all):
hseqlist = psl.get('tseqs')
hseq = '' if not hseqlist else hseqlist[idx]
qseqlist = psl.get('qseqs')
qseq = '' if not qseqlist else qseqlist[idx]
frag = HSPFragment(hid, qid, hit=hseq, query=qseq)
# set alphabet
frag.alphabet = generic_dna
# set coordinates
frag.query_start = qcoords[0]
frag.query_end = qcoords[1]
frag.hit_start = hit_range_all[idx][0]
frag.hit_end = hit_range_all[idx][1]
# and strands
frag.query_strand = qstrand
frag.hit_strand = hstrand
frags.append(frag)
# create hsp object
hsp = HSP(frags)
# check if start and end are set correctly
assert hsp.query_start == psl['qstart']
assert hsp.query_end == psl['qend']
assert hsp.hit_start == psl['tstart']
assert hsp.hit_end == psl['tend']
# and check block spans as well
assert hsp.query_span_all == hsp.hit_span_all == psl['blocksizes']
# set its attributes
hsp.match_num = psl['matches']
hsp.mismatch_num = psl['mismatches']
hsp.match_rep_num = psl['repmatches']
hsp.n_num = psl['ncount']
hsp.query_gapopen_num = psl['qnuminsert']
hsp.query_gap_num = psl['qbaseinsert']
hsp.hit_gapopen_num = psl['tnuminsert']
hsp.hit_gap_num = psl['tbaseinsert']
hsp.ident_num = psl['matches'] + psl['repmatches']
hsp.gapopen_num = psl['qnuminsert'] + psl['tnuminsert']
hsp.gap_num = psl['qbaseinsert'] + psl['tbaseinsert']
hsp.query_is_protein = is_protein
hsp.ident_pct = 100.0 - _calc_millibad(psl, is_protein) * 0.1
hsp.score = _calc_score(psl, is_protein)
# helper flag, for writing
hsp._has_hit_strand = len(psl['strand']) == 2
return hsp
def parse_vulgar_comp(hsp, vulgar_comp):
"""Parses the vulgar components present in the hsp dictionary."""
# containers for block coordinates
qstarts, qends, hstarts, hends = \
[hsp['query_start']], [], [hsp['hit_start']], []
# containers for split codons
hsp['query_split_codons'], hsp['hit_split_codons'] = [], []
# containers for ner blocks
hsp['query_ner_ranges'], hsp['hit_ner_ranges'] = [], []
# sentinels for tracking query and hit positions
qpos, hpos = hsp['query_start'], hsp['hit_start']
# multiplier for determining sentinel movement
qmove = 1 if hsp['query_strand'] >= 0 else -1
hmove = 1 if hsp['hit_strand'] >= 0 else -1
vcomps = re.findall(_RE_VCOMP, vulgar_comp)
for idx, match in enumerate(vcomps):
label, qstep, hstep = match[0], int(match[1]), int(match[2])
# check for label, must be recognized
assert label in 'MCGF53INS', "Unexpected vulgar label: %r" % label
# match, codon, or gaps
if label in 'MCGS':
# if the previous comp is not an MCGS block, it's the
# start of a new block
if vcomps[idx-1][0] not in 'MCGS':
qstarts.append(qpos)
hstarts.append(hpos)
# other labels
# store the values in the hsp dict as a tuple of (start, stop)
# we're not doing anything if the label is in '53IN', as these
# basically tell us what the inter-block coordinates are and
# inter-block coordinates are automatically calculated by
# and HSP property
if label == 'S':
# get start and stop from parsed values
qstart, hstart = qpos, hpos
qend = qstart + qstep * qmove
hend = hstart + hstep * hmove
# adjust the start-stop ranges
sqstart, sqend = min(qstart, qend), max(qstart, qend)
shstart, shend = min(hstart, hend), max(hstart, hend)
# split codons
# XXX: is it possible to have a frameshift that introduces
# a codon split? If so, this may need a different treatment..
hsp['query_split_codons'].append((sqstart, sqend))
hsp['hit_split_codons'].append((shstart, shend))
# move sentinels accordingly
qpos += qstep * qmove
hpos += hstep * hmove
# append to ends if the next comp is not an MCGS block or
# if it's the last comp
if idx == len(vcomps)-1 or \
(label in 'MCGS' and vcomps[idx+1][0] not in 'MCGS'):
qends.append(qpos)
hends.append(hpos)
# adjust coordinates
for seq_type in ('query_', 'hit_'):
strand = hsp[seq_type + 'strand']
# switch coordinates if strand is < 0
if strand < 0:
# switch the starts and ends
hsp[seq_type + 'start'], hsp[seq_type + 'end'] = \
hsp[seq_type + 'end'], hsp[seq_type + 'start']
if seq_type == 'query_':
qstarts, qends = qends, qstarts
else:
hstarts, hends = hends, hstarts
# set start and end ranges
hsp['query_ranges'] = list(zip(qstarts, qends))
hsp['hit_ranges'] = list(zip(hstarts, hends))
return hsp