def getCopy(self):
"""return a new copy.
"""
new_entry = Prediction()
new_entry.mExpand = self.mExpand
new_entry.mPredictionId = self.mPredictionId
new_entry.mQueryToken = self.mQueryToken
new_entry.mQueryFrom = self.mQueryFrom
new_entry.mQueryTo = self.mQueryTo
new_entry.mSbjctToken = self.mSbjctToken
new_entry.mSbjctStrand = self.mSbjctStrand
new_entry.mSbjctFrom = self.mSbjctFrom
new_entry.mSbjctTo = self.mSbjctTo
new_entry.mRank = self.mRank
new_entry.score = self.score
new_entry.mQueryLength = self.mQueryLength
new_entry.mQueryCoverage = self.mQueryCoverage
new_entry.mNGaps = self.mNGaps
new_entry.mNFrameShifts = self.mNFrameShifts
new_entry.mNIntrons = self.mNIntrons
new_entry.mNSplits = self.mNSplits
new_entry.mNStopCodons = self.mNStopCodons
new_entry.mPercentIdentity = self.mPercentIdentity
new_entry.mPercentSimilarity = self.mPercentSimilarity
new_entry.mTranslation = self.mTranslation
new_entry.mSbjctGenomeFrom = self.mSbjctGenomeFrom
new_entry.mSbjctGenomeTo = self.mSbjctGenomeTo
new_entry.mAlignmentString = self.mAlignmentString
new_entry.mQueryAli = self.mQueryAli
new_entry.mSbjctAli = self.mSbjctAli
if self.mExpand:
new_entry.mMapPeptide2Translation = alignlib_lite.py_makeAlignmentVector(
)
alignlib_lite.py_copyAlignment(new_entry.mMapPeptide2Translation,
self.mMapPeptide2Translation)
new_entry.mMapPeptide2Genome = Genomics.String2Alignment(
new_entry.mAlignmentString)
else:
new_entry.mMapPeptide2Translation = self.mMapPeptide2Translation = None
new_entry.mMapPeptide2Genome = self.mMapPeptide2Genome = None
return new_entry
def getCopy( self ):
"""return a new copy.
"""
new_entry = Prediction()
new_entry.mExpand = self.mExpand
new_entry.mPredictionId = self.mPredictionId
new_entry.mQueryToken = self.mQueryToken
new_entry.mQueryFrom = self.mQueryFrom
new_entry.mQueryTo = self.mQueryTo
new_entry.mSbjctToken = self.mSbjctToken
new_entry.mSbjctStrand = self.mSbjctStrand
new_entry.mSbjctFrom = self.mSbjctFrom
new_entry.mSbjctTo = self.mSbjctTo
new_entry.mRank = self.mRank
new_entry.score = self.score
new_entry.mQueryLength = self.mQueryLength
new_entry.mQueryCoverage = self.mQueryCoverage
new_entry.mNGaps = self.mNGaps
new_entry.mNFrameShifts = self.mNFrameShifts
new_entry.mNIntrons = self.mNIntrons
new_entry.mNSplits = self.mNSplits
new_entry.mNStopCodons = self.mNStopCodons
new_entry.mPercentIdentity = self.mPercentIdentity
new_entry.mPercentSimilarity = self.mPercentSimilarity
new_entry.mTranslation = self.mTranslation
new_entry.mSbjctGenomeFrom = self.mSbjctGenomeFrom
new_entry.mSbjctGenomeTo = self.mSbjctGenomeTo
new_entry.mAlignmentString = self.mAlignmentString
new_entry.mQueryAli = self.mQueryAli
new_entry.mSbjctAli = self.mSbjctAli
if self.mExpand:
new_entry.mMapPeptide2Translation = alignlib_lite.py_makeAlignmentVector()
alignlib_lite.py_copyAlignment( new_entry.mMapPeptide2Translation, self.mMapPeptide2Translation)
new_entry.mMapPeptide2Genome = Genomics.String2Alignment( new_entry.mAlignmentString)
else:
new_entry.mMapPeptide2Translation = self.mMapPeptide2Translation = None
new_entry.mMapPeptide2Genome = self.mMapPeptide2Genome = None
return new_entry
def main(argv=None):
"""script main.
parses command line options in sys.argv, unless *argv* is given.
"""
if argv is None:
argv = sys.argv
parser = E.OptionParser(
version=
"%prog version: $Id: links2fasta.py 2446 2009-01-27 16:32:35Z andreas $",
usage=globals()["__doc__"])
parser.add_option("-s",
"--sequences",
dest="filename_sequences",
type="string",
help="peptide sequence [Default=%default]")
parser.add_option("-f",
"--format",
dest="format",
type="string",
help="output format [Default=%default]")
parser.add_option(
"-e",
"--expand",
dest="expand",
action="store_true",
help=
"expand positions from peptide to nucleotide alignment [Default=%default]"
)
parser.add_option("-m",
"--map",
dest="filename_map",
type="string",
help="map alignments [Default=%default]")
parser.add_option("-c",
"--codons",
dest="require_codons",
action="store_true",
help="require codons [Default=%default]")
parser.add_option(
"--one-based-coordinates",
dest="one_based_coordinates",
action="store_true",
help=
"expect one-based coordinates. The default are zero based coordinates [Default=%default]."
)
parser.add_option("--no-identical",
dest="no_identical",
action="store_true",
help="do not output identical pairs [Default=%default]")
parser.add_option(
"-g",
"--no-gaps",
dest="no_gaps",
action="store_true",
help="remove all gaps from aligned sequences [Default=%default]")
parser.add_option("-x",
"--exons",
dest="filename_exons",
type="string",
help="filename with exon boundaries [Default=%default]")
parser.add_option("-o",
"--outfile",
dest="filename_outfile",
type="string",
help="filename to save links [Default=%default]")
parser.add_option("--min-length",
dest="min_length",
type="int",
help="minimum length of alignment [Default=%default]")
parser.add_option(
"--filter",
dest="filename_filter",
type="string",
help=
"given a set of previous alignments, only write new pairs [Default=%default]."
)
parser.set_defaults(filename_sequences=None,
filename_exons=None,
filename_map=None,
filename_outfile=None,
no_gaps=False,
format="fasta",
expand=False,
require_codons=False,
no_identical=False,
min_length=0,
report_step=100,
one_based_coordinates=False,
filename_filter=None)
(options, args) = E.Start(parser, add_mysql_options=True)
t0 = time.time()
if options.filename_sequences:
sequences = Genomics.ReadPeptideSequences(
open(options.filename_sequences, "r"))
else:
sequences = {}
if options.loglevel >= 1:
options.stdlog.write("# read %i sequences\n" % len(sequences))
sys.stdout.flush()
if options.filename_exons:
exons = Exons.ReadExonBoundaries(open(options.filename_exons, "r"))
else:
exons = {}
if options.loglevel >= 1:
options.stdlog.write("# read %i exons\n" % len(exons))
sys.stdout.flush()
if options.filename_map:
map_old2new = {}
for line in open(options.filename_map, "r"):
if line[0] == "#":
continue
m = Map()
m.read(line)
map_old2new[m.mToken] = m
else:
map_old2new = {}
if options.loglevel >= 1:
options.stdlog.write("# read %i maps\n" % len(map_old2new))
sys.stdout.flush()
if options.filename_filter:
if options.loglevel >= 1:
options.stdlog.write("# reading filtering information.\n")
sys.stdout.flush()
map_pair2hids = {}
if os.path.exists(options.filename_filter):
infile = open(options.filename_filter, "r")
iterator = FastaIterator.FastaIterator(infile)
while 1:
cur_record = iterator.next()
if cur_record is None:
break
record1 = cur_record
cur_record = iterator.next()
if cur_record is None:
break
record2 = cur_record
identifier1 = re.match("(\S+)", record1.title).groups()[0]
identifier2 = re.match("(\S+)", record2.title).groups()[0]
id = "%s-%s" % (identifier1, identifier2)
s = Genomics.GetHID(record1.sequence + ";" + record2.sequence)
if id not in map_pair2hids:
map_pair2hids[id] = []
map_pair2hids[id].append(s)
infile.close()
if options.loglevel >= 1:
options.stdlog.write(
"# read filtering information for %i pairs.\n" %
len(map_pair2hids))
sys.stdout.flush()
else:
map_pair2hids = None
if options.loglevel >= 1:
options.stdlog.write("# finished input in %i seconds.\n" %
(time.time() - t0))
if options.filename_outfile:
outfile = open(options.filename_outfile, "w")
else:
outfile = None
map_row2col = alignlib_lite.py_makeAlignmentVector()
tmp1_map_row2col = alignlib_lite.py_makeAlignmentVector()
counts = {}
iterations = 0
t1 = time.time()
ninput, nskipped, noutput = 0, 0, 0
for link in BlastAlignments.iterator_links(sys.stdin):
iterations += 1
ninput += 1
if options.loglevel >= 1:
if (iterations % options.report_step == 0):
options.stdlog.write("# iterations: %i in %i seconds.\n" %
(iterations, time.time() - t1))
sys.stdout.flush()
if link.mQueryToken not in sequences or \
link.mSbjctToken not in sequences:
nskipped += 1
continue
if options.loglevel >= 3:
options.stdlog.write("# read link %s\n" % str(link))
row_seq = alignlib_lite.py_makeSequence(sequences[link.mQueryToken])
col_seq = alignlib_lite.py_makeSequence(sequences[link.mSbjctToken])
if options.one_based_coordinates:
link.mQueryFrom -= 1
link.mSbjctFrom -= 1
if options.expand:
link.mQueryFrom = link.mQueryFrom * 3
link.mSbjctFrom = link.mSbjctFrom * 3
link.mQueryAli = ScaleAlignment(link.mQueryAli, 3)
link.mSbjctAli = ScaleAlignment(link.mSbjctAli, 3)
map_row2col.clear()
alignlib_lite.py_AlignmentFormatEmissions(
link.mQueryFrom, link.mQueryAli, link.mSbjctFrom,
link.mSbjctAli).copy(map_row2col)
if link.mQueryToken in map_old2new:
tmp1_map_row2col.clear()
map_old2new[link.mQueryToken].expand()
if options.loglevel >= 3:
options.stdlog.write("# combining in row with %s\n" % str(
alignlib_lite.py_AlignmentFormatEmissions(
map_old2new[link.mQueryToken].mMapOld2New)))
alignlib_lite.py_combineAlignment(
tmp1_map_row2col, map_old2new[link.mQueryToken].mMapOld2New,
map_row2col, alignlib_lite.py_RR)
map_old2new[link.mQueryToken].clear()
alignlib_lite.py_copyAlignment(map_row2col, tmp1_map_row2col)
if link.mSbjctToken in map_old2new:
tmp1_map_row2col.clear()
map_old2new[link.mSbjctToken].expand()
if options.loglevel >= 3:
options.stdlog.write("# combining in col with %s\n" % str(
alignlib_lite.py_AlignmentFormatEmissions(
map_old2new[link.mSbjctToken].mMapOld2New)))
alignlib_lite.py_combineAlignment(
tmp1_map_row2col, map_row2col,
map_old2new[link.mSbjctToken].mMapOld2New, alignlib_lite.py_CR)
map_old2new[link.mSbjctToken].clear()
alignlib_lite.py_copyAlignment(map_row2col, tmp1_map_row2col)
dr = row_seq.getLength() - map_row2col.getRowTo()
dc = col_seq.getLength() - map_row2col.getColTo()
if dr < 0 or dc < 0:
raise ValueError(
"out of bounds alignment: %s-%s: alignment out of bounds. row=%i col=%i ali=%s"
%
(link.mQueryToken, link.mSbjctToken, row_seq.getLength(),
col_seq.getLength(),
str(alignlib_lite.py_AlignmentFormatEmissions(map_row2col))))
if options.loglevel >= 2:
options.stdlog.write(
str(
alignlib_lite.py_AlignmentFormatExplicit(
map_row2col, row_seq, col_seq)) + "\n")
# check for incomplete codons
if options.require_codons:
naligned = map_row2col.getNumAligned()
# turned off, while fixing alignlib_lite
if naligned % 3 != 0:
options.stdlog.write("# %s\n" % str(map_row2col))
options.stdlog.write("# %s\n" % str(link))
options.stdlog.write("# %s\n" %
str(map_old2new[link.mQueryToken]))
options.stdlog.write("# %s\n" %
str(map_old2new[link.mSbjctToken]))
options.stdlog.write("#\n%s\n" %
alignlib_lite.py_AlignmentFormatExplicit(
map_row2col, row_seq, col_seq))
raise ValueError(
"incomplete codons %i in pair %s - %s" %
(naligned, link.mQueryToken, link.mSbjctToken))
# if so desired, write on a per exon level:
if exons:
if link.mQueryToken not in exons:
raise IndexError("%s not found in exons" % (link.mQueryToken))
if link.mSbjctToken not in exons:
raise IndexError("%s not found in exons" % (link.mSbjctToken))
exons1 = exons[link.mQueryToken]
exons2 = exons[link.mSbjctToken]
# Get overlapping segments
segments = Exons.MatchExons(map_row2col, exons1, exons2)
for a, b in segments:
tmp1_map_row2col.clear()
# make sure you got codon boundaries. Note that frameshifts
# in previous exons will cause the codons to start at positions
# different from mod 3. The problem is that I don't know where
# the frameshifts occur exactly. The exon boundaries are given
# with respect to the cds, which include the frame shifts.
# Unfortunately, phase information seems to be incomplete in
# the input files.
from1, to1 = GetAdjustedBoundaries(a, exons1)
from2, to2 = GetAdjustedBoundaries(b, exons2)
alignlib_lite.py_copyAlignment(tmp1_map_row2col, map_row2col,
from1 + 1, to1, from2 + 1, to2)
mode = Write(tmp1_map_row2col,
row_seq,
col_seq,
link,
no_gaps=options.no_gaps,
no_identical=options.no_identical,
min_length=options.min_length,
suffix1="_%s" % str(a),
suffix2="_%s" % str(b),
outfile=outfile,
pair_filter=map_pair2hid,
format=options.format)
if mode not in counts:
counts[mode] = 0
counts[mode] += 1
else:
mode = Write(map_row2col,
row_seq,
col_seq,
link,
min_length=options.min_length,
no_gaps=options.no_gaps,
no_identical=options.no_identical,
outfile=outfile,
pair_filter=map_pair2hids,
format=options.format)
if mode not in counts:
counts[mode] = 0
counts[mode] += 1
noutput += 1
if outfile:
outfile.close()
if options.loglevel >= 1:
options.stdlog.write("# %s\n" % ", ".join(
map(lambda x, y: "%s=%i" %
(x, y), counts.keys(), counts.values())))
options.stdlog.write("# ninput=%i, noutput=%i, nskipped=%i\n" %
(ninput, noutput, nskipped))
E.Stop()
def pslMap(options):
"""thread psl alignments using intervals.
"""
if options.format == "gtf":
use_copy = False
else:
use_copy = True
c = E.Counter()
min_length = options.min_aligned
for match, qx, tx in iterator_psl_intervals(options):
map_query2target = match.getMapQuery2Target()
c.input += 1
# if no filter on qx or tx, use full segment
if qx is None:
qx = [(match.mQueryFrom, match.mQueryTo, 0)]
elif tx is None:
tx = [(match.mSbjctFrom, match.mSbjctTo, 0)]
E.debug('matches in query: %s' % qx)
E.debug('matches in target: %s' % tx)
# if no overlap: return
if not qx or not tx:
c.skipped += 1
E.debug("no matches in query or target - skipped")
continue
for query in qx:
qstart, qend, qval = query
# skip elements that are too small
if qend - qstart < min_length:
E.debug("query too small - skipped at %s:%i-%i" %
(match.mQueryId, qstart, qend))
c.skipped_small_queries += 1
continue
E.debug("working on query %s:%i-%i" %
(match.mQueryId, qstart, qend))
mqstart, mqend = (
map_query2target.mapRowToCol(
qstart,
alignlib_lite.py_RIGHT),
map_query2target.mapRowToCol(
qend,
alignlib_lite.py_LEFT))
if match.strand == "-":
qstart, qend = match.mQueryLength - \
qend, match.mQueryLength - qstart
for target in tx:
tstart, tend, tval = target
if (tstart >= mqend or tend <= mqstart):
E.debug("no overlap: %i-%i (%i-%i) - %i-%i" % (
qstart, qend, mqstart, mqend, tstart, tend))
continue
if tend - tstart < min_length:
E.debug("target length too short: %i-%i - %i-%i" % (
qstart, qend, tstart, tend))
continue
new = alignlib_lite.py_makeAlignmentBlocks()
if use_copy:
# do copy with range filter
if options.loglevel >= 3:
mtstart, mtend = map_query2target.mapColToRow(
tstart), map_query2target.mapColToRow(tend)
E.debug(
("query: %i-%i (len=%i)-> %i-%i(len=%i); "
"target: %i-%i (len=%i)-> %i-%i (len=%i)") %
(qstart, qend,
qend - qstart,
mqstart, mqend,
mqend - mqstart,
tstart, tend,
tend - tstart,
mtstart, mtend,
mtend - mtstart))
alignlib_lite.py_copyAlignment(
new,
map_query2target,
qstart, qend,
tstart, tend)
else:
# do copy with alignment filter
map_query = qval
if map_query:
tmp = alignlib_lite.py_makeAlignmentBlocks()
alignlib_lite.py_copyAlignment(
tmp, map_query2target, map_query,
alignlib_lite.py_RR)
if options.loglevel >= 5:
options.stdlog.write(
"######## mapping query ###########\n")
options.stdlog.write(
"# %s\n" %
str(alignlib_lite.py_AlignmentFormatEmissions(
map_query2target)))
options.stdlog.write(
"# %s\n" % str(
alignlib_lite.py_AlignmentFormatEmissions(
map_query)))
options.stdlog.write(
"# %s\n" % str(
alignlib_lite.py_AlignmentFormatEmissions(
tmp)))
else:
tmp = map_query2target
map_target = tval
if map_target:
new = alignlib_lite.py_makeAlignmentBlocks()
alignlib_lite.py_copyAlignment(
new, tmp, map_target, alignlib_lite.py_CR)
if options.loglevel >= 5:
options.stdlog.write(
"######## mapping target ###########\n")
options.stdlog.write(
"# before: %s\n" %
str(alignlib_lite.py_AlignmentFormatEmissions(
tmp)))
options.stdlog.write(
"# map : %s\n" %
str(alignlib_lite.py_AlignmentFormatEmissions(
map_target)))
options.stdlog.write(
"# after : %s\n" %
str(alignlib_lite.py_AlignmentFormatEmissions(
new)))
else:
new = tmp
if options.loglevel >= 4:
E.debug("putative match with intervals: %s and %s: %i-%i" %
(str(query), str(target), qstart, qend))
if options.loglevel >= 5:
E.debug(
"input : %s" % str(
alignlib_lite.py_AlignmentFormatEmissions(
map_query2target)))
E.debug("final : %s" %
str(alignlib_lite.py_AlignmentFormatEmissions(
new)))
if new.getLength() > 0:
n = match.copy()
n.fromMap(new, use_strand=True)
E.info("match : %s" % (str(n)))
if new.getNumAligned() > options.min_aligned:
n = match.copy()
n.fromMap(new, use_strand=True)
options.stdout.write(str(n) + "\n")
c.output += 1
else:
c.discarded += 1
break
else:
c.nooverlap += 1
E.info("map: %s" % str(c))
def pslMap(options):
"""thread psl alignments using intervals.
"""
if options.format == "gtf":
use_copy = False
else:
use_copy = True
c = E.Counter()
min_length = options.min_aligned
for match, qx, tx in iterator_psl_intervals(options):
map_query2target = match.getMapQuery2Target()
c.input += 1
# if no filter on qx or tx, use full segment
if qx is None:
qx = [(match.mQueryFrom, match.mQueryTo, 0)]
elif tx is None:
tx = [(match.mSbjctFrom, match.mSbjctTo, 0)]
E.debug('matches in query: %s' % qx)
E.debug('matches in target: %s' % tx)
# if no overlap: return
if not qx or not tx:
c.skipped += 1
E.debug("no matches in query or target - skipped")
continue
for query in qx:
qstart, qend, qval = query
# skip elements that are too small
if qend - qstart < min_length:
E.debug("query too small - skipped at %s:%i-%i" %
(match.mQueryId, qstart, qend))
c.skipped_small_queries += 1
continue
E.debug("working on query %s:%i-%i" %
(match.mQueryId, qstart, qend))
mqstart, mqend = (
map_query2target.mapRowToCol(
qstart,
alignlib_lite.py_RIGHT),
map_query2target.mapRowToCol(
qend,
alignlib_lite.py_LEFT))
if match.strand == "-":
qstart, qend = match.mQueryLength - \
qend, match.mQueryLength - qstart
for target in tx:
tstart, tend, tval = target
if (tstart >= mqend or tend <= mqstart):
E.debug("no overlap: %i-%i (%i-%i) - %i-%i" % (
qstart, qend, mqstart, mqend, tstart, tend))
continue
if tend - tstart < min_length:
E.debug("target length too short: %i-%i - %i-%i" % (
qstart, qend, tstart, tend))
continue
new = alignlib_lite.py_makeAlignmentBlocks()
if use_copy:
# do copy with range filter
if options.loglevel >= 3:
mtstart, mtend = map_query2target.mapColToRow(
tstart), map_query2target.mapColToRow(tend)
E.debug(
("query: %i-%i (len=%i)-> %i-%i(len=%i); "
"target: %i-%i (len=%i)-> %i-%i (len=%i)") %
(qstart, qend,
qend - qstart,
mqstart, mqend,
mqend - mqstart,
tstart, tend,
tend - tstart,
mtstart, mtend,
mtend - mtstart))
alignlib_lite.py_copyAlignment(
new,
map_query2target,
qstart, qend,
tstart, tend)
else:
# do copy with alignment filter
map_query = qval
if map_query:
tmp = alignlib_lite.py_makeAlignmentBlocks()
alignlib_lite.py_copyAlignment(
tmp, map_query2target, map_query,
alignlib_lite.py_RR)
if options.loglevel >= 5:
options.stdlog.write(
"######## mapping query ###########\n")
options.stdlog.write(
"# %s\n" %
str(alignlib_lite.py_AlignmentFormatEmissions(
map_query2target)))
options.stdlog.write(
"# %s\n" % str(
alignlib_lite.py_AlignmentFormatEmissions(
map_query)))
options.stdlog.write(
"# %s\n" % str(
alignlib_lite.py_AlignmentFormatEmissions(
tmp)))
else:
tmp = map_query2target
map_target = tval
if map_target:
new = alignlib_lite.py_makeAlignmentBlocks()
alignlib_lite.py_copyAlignment(
new, tmp, map_target, alignlib_lite.py_CR)
if options.loglevel >= 5:
options.stdlog.write(
"######## mapping target ###########\n")
options.stdlog.write(
"# before: %s\n" %
str(alignlib_lite.py_AlignmentFormatEmissions(
tmp)))
options.stdlog.write(
"# map : %s\n" %
str(alignlib_lite.py_AlignmentFormatEmissions(
map_target)))
options.stdlog.write(
"# after : %s\n" %
str(alignlib_lite.py_AlignmentFormatEmissions(
new)))
else:
new = tmp
if options.loglevel >= 4:
E.debug("putative match with intervals: %s and %s: %i-%i" %
(str(query), str(target), qstart, qend))
if options.loglevel >= 5:
E.debug(
"input : %s" % str(
alignlib_lite.py_AlignmentFormatEmissions(
map_query2target)))
E.debug("final : %s" %
str(alignlib_lite.py_AlignmentFormatEmissions(
new)))
if new.getLength() > 0:
n = match.copy()
n.fromMap(new, use_strand=True)
E.info("match : %s" % (str(n)))
if new.getNumAligned() > options.min_aligned:
n = match.copy()
n.fromMap(new, use_strand=True)
options.stdout.write(str(n) + "\n")
c.output += 1
else:
c.discarded += 1
break
else:
c.nooverlap += 1
E.info("map: %s" % str(c))
def main( argv = None ):
"""script main.
parses command line options in sys.argv, unless *argv* is given.
"""
if argv == None: argv = sys.argv
parser = E.OptionParser( version = "%prog version: $Id: links2fasta.py 2446 2009-01-27 16:32:35Z andreas $", usage = globals()["__doc__"] )
parser.add_option( "-s", "--sequences", dest="filename_sequences", type="string",
help="peptide sequence [Default=%default]" )
parser.add_option( "-f", "--format", dest="format", type="string",
help="output format [Default=%default]" )
parser.add_option( "-e", "--expand", dest="expand", action="store_true",
help="expand positions from peptide to nucleotide alignment [Default=%default]")
parser.add_option( "-m", "--map", dest="filename_map", type="string",
help="map alignments [Default=%default]")
parser.add_option( "-c", "--codons", dest="require_codons", action="store_true",
help="require codons [Default=%default]")
parser.add_option( "--one-based-coordinates", dest="one_based_coordinates", action="store_true",
help="expect one-based coordinates. The default are zero based coordinates [Default=%default].")
parser.add_option( "--no-identical", dest="no_identical", action="store_true",
help="do not output identical pairs [Default=%default]" )
parser.add_option( "-g", "--no-gaps", dest="no_gaps", action="store_true",
help="remove all gaps from aligned sequences [Default=%default]")
parser.add_option( "-x", "--exons", dest="filename_exons", type="string",
help="filename with exon boundaries [Default=%default]")
parser.add_option( "-o", "--outfile", dest="filename_outfile", type="string",
help="filename to save links [Default=%default]")
parser.add_option( "--min-length", dest="min_length", type="int",
help="minimum length of alignment [Default=%default]")
parser.add_option( "--filter", dest="filename_filter", type="string",
help="given a set of previous alignments, only write new pairs [Default=%default].")
parser.set_defaults(
filename_sequences = None,
filename_exons = None,
filename_map = None,
filename_outfile = None,
no_gaps = False,
format = "fasta",
expand = False,
require_codons = False,
no_identical = False,
min_length = 0,
report_step = 100,
one_based_coordinates = False,
filename_filter = None)
(options, args) = E.Start( parser, add_mysql_options = True )
t0 = time.time()
if options.filename_sequences:
sequences = Genomics.ReadPeptideSequences( open(options.filename_sequences, "r") )
else:
sequences = {}
if options.loglevel >= 1:
options.stdlog.write( "# read %i sequences\n" % len(sequences) )
sys.stdout.flush()
if options.filename_exons:
exons = Exons.ReadExonBoundaries( open(options.filename_exons, "r") )
else:
exons = {}
if options.loglevel >= 1:
options.stdlog.write( "# read %i exons\n" % len(exons) )
sys.stdout.flush()
if options.filename_map:
map_old2new = {}
for line in open(options.filename_map, "r"):
if line[0] == "#": continue
m = Map()
m.read( line )
map_old2new[m.mToken] = m
else:
map_old2new = {}
if options.loglevel >= 1:
options.stdlog.write( "# read %i maps\n" % len(map_old2new) )
sys.stdout.flush()
if options.filename_filter:
if options.loglevel >= 1:
options.stdlog.write( "# reading filtering information.\n" )
sys.stdout.flush()
map_pair2hids = {}
if os.path.exists( options.filename_filter ):
infile = open(options.filename_filter, "r")
iterator = FastaIterator.FastaIterator( infile )
while 1:
cur_record = iterator.next()
if cur_record is None: break
record1 = cur_record
cur_record = iterator.next()
if cur_record is None: break
record2 = cur_record
identifier1 = re.match("(\S+)", record1.title).groups()[0]
identifier2 = re.match("(\S+)", record2.title).groups()[0]
id = "%s-%s" % (identifier1, identifier2)
s = Genomics.GetHID(record1.sequence + ";" + record2.sequence)
if id not in map_pair2hids: map_pair2hids[id] = []
map_pair2hids[id].append( s )
infile.close()
if options.loglevel >= 1:
options.stdlog.write( "# read filtering information for %i pairs.\n" % len(map_pair2hids) )
sys.stdout.flush()
else:
map_pair2hids = None
if options.loglevel >= 1:
options.stdlog.write( "# finished input in %i seconds.\n" % (time.time() - t0))
if options.filename_outfile:
outfile = open(options.filename_outfile, "w")
else:
outfile = None
map_row2col = alignlib_lite.py_makeAlignmentVector()
tmp1_map_row2col = alignlib_lite.py_makeAlignmentVector()
counts = {}
iterations = 0
t1 = time.time()
ninput, nskipped, noutput = 0, 0, 0
for link in BlastAlignments.iterator_links( sys.stdin ):
iterations += 1
ninput += 1
if options.loglevel >= 1:
if (iterations % options.report_step == 0):
options.stdlog.write( "# iterations: %i in %i seconds.\n" % (iterations, time.time() - t1) )
sys.stdout.flush()
if link.mQueryToken not in sequences or \
link.mSbjctToken not in sequences:
nskipped += 1
continue
if options.loglevel >= 3:
options.stdlog.write( "# read link %s\n" % str(link) )
row_seq = alignlib_lite.py_makeSequence( sequences[link.mQueryToken] )
col_seq = alignlib_lite.py_makeSequence( sequences[link.mSbjctToken] )
if options.one_based_coordinates:
link.mQueryFrom -= 1
link.mSbjctFrom -= 1
if options.expand:
link.mQueryFrom = link.mQueryFrom * 3
link.mSbjctFrom = link.mSbjctFrom * 3
link.mQueryAli = ScaleAlignment( link.mQueryAli, 3 )
link.mSbjctAli = ScaleAlignment( link.mSbjctAli, 3 )
map_row2col.clear()
alignlib_lite.py_AlignmentFormatEmissions(
link.mQueryFrom, link.mQueryAli,
link.mSbjctFrom, link.mSbjctAli ).copy( map_row2col )
if link.mQueryToken in map_old2new:
tmp1_map_row2col.clear()
map_old2new[link.mQueryToken].expand()
if options.loglevel >= 3:
options.stdlog.write( "# combining in row with %s\n" %\
str(alignlib_lite.py_AlignmentFormatEmissions(map_old2new[link.mQueryToken].mMapOld2New ) ))
alignlib_lite.py_combineAlignment( tmp1_map_row2col,
map_old2new[link.mQueryToken].mMapOld2New,
map_row2col,
alignlib_lite.py_RR )
map_old2new[link.mQueryToken].clear()
alignlib_lite.py_copyAlignment( map_row2col, tmp1_map_row2col )
if link.mSbjctToken in map_old2new:
tmp1_map_row2col.clear()
map_old2new[link.mSbjctToken].expand()
if options.loglevel >= 3:
options.stdlog.write( "# combining in col with %s\n" %\
str(alignlib_lite.py_AlignmentFormatEmissions(map_old2new[link.mSbjctToken].mMapOld2New ) ))
alignlib_lite.py_combineAlignment( tmp1_map_row2col,
map_row2col,
map_old2new[link.mSbjctToken].mMapOld2New,
alignlib_lite.py_CR )
map_old2new[link.mSbjctToken].clear()
alignlib_lite.py_copyAlignment( map_row2col, tmp1_map_row2col )
dr = row_seq.getLength() - map_row2col.getRowTo()
dc = col_seq.getLength() - map_row2col.getColTo()
if dr < 0 or dc < 0:
raise ValueError("out of bounds alignment: %s-%s: alignment out of bounds. row=%i col=%i ali=%s" %\
(link.mQueryToken,
link.mSbjctToken,
row_seq.getLength(),
col_seq.getLength(),
str(alignlib_lite.py_AlignmentFormatEmissions(map_row2col))))
if options.loglevel >= 2:
options.stdlog.write( str( alignlib_lite.py_AlignmentFormatExplicit( map_row2col,
row_seq,
col_seq )) + "\n" )
## check for incomplete codons
if options.require_codons:
naligned = map_row2col.getNumAligned()
# turned off, while fixing alignlib_lite
if naligned % 3 != 0:
options.stdlog.write( "# %s\n" % str(map_row2col) )
options.stdlog.write( "# %s\n" % str(link) )
options.stdlog.write( "# %s\n" % str(map_old2new[link.mQueryToken]) )
options.stdlog.write( "# %s\n" % str(map_old2new[link.mSbjctToken]) )
options.stdlog.write( "#\n%s\n" % alignlib_lite.py_AlignmentFormatExplicit( map_row2col,
row_seq,
col_seq ) )
raise ValueError("incomplete codons %i in pair %s - %s" % (naligned, link.mQueryToken, link.mSbjctToken))
## if so desired, write on a per exon level:
if exons:
if link.mQueryToken not in exons:
raise IndexError("%s not found in exons" % (link.mQueryToken))
if link.mSbjctToken not in exons:
raise IndexError("%s not found in exons" % (link.mSbjctToken))
exons1 = exons[link.mQueryToken]
exons2 = exons[link.mSbjctToken]
## Get overlapping segments
segments = Exons.MatchExons( map_row2col, exons1, exons2 )
for a,b in segments:
tmp1_map_row2col.clear()
# make sure you got codon boundaries. Note that frameshifts
# in previous exons will cause the codons to start at positions
# different from mod 3. The problem is that I don't know where
# the frameshifts occur exactly. The exon boundaries are given
# with respect to the cds, which include the frame shifts.
# Unfortunately, phase information seems to be incomplete in the input files.
from1, to1 = GetAdjustedBoundaries( a, exons1 )
from2, to2 = GetAdjustedBoundaries( b, exons2 )
alignlib_lite.py_copyAlignment( tmp1_map_row2col, map_row2col,
from1+1, to1, from2+1, to2 )
mode = Write( tmp1_map_row2col, row_seq, col_seq, link,
no_gaps = options.no_gaps,
no_identical = options.no_identical,
min_length = options.min_length,
suffix1="_%s" % str(a),
suffix2="_%s" % str(b),
outfile = outfile,
pair_filter = map_pair2hid,
format = options.format )
if mode not in counts: counts[mode] = 0
counts[mode] += 1
else:
mode = Write( map_row2col, row_seq, col_seq, link,
min_length = options.min_length,
no_gaps = options.no_gaps,
no_identical = options.no_identical,
outfile = outfile,
pair_filter = map_pair2hids,
format = options.format )
if mode not in counts: counts[mode] = 0
counts[mode] += 1
noutput += 1
if outfile: outfile.close()
if options.loglevel >= 1:
options.stdlog.write("# %s\n" % ", ".join( map( lambda x,y: "%s=%i" % (x,y), counts.keys(), counts.values() ) ))
options.stdlog.write("# ninput=%i, noutput=%i, nskipped=%i\n" % (ninput, noutput, nskipped) )
E.Stop()
