def GetMapSequences(sequences):
map_cluster2sequence = {}
map_sequence2cluster = {}
# cluster sequences by identity
# (clumsy sort, use hashes for bigger sets)
for key, sequence in sequences.items():
h = Genomics.GetHID(sequence)
if h not in map_cluster2sequence:
map_cluster2sequence[h] = []
map_sequence2cluster[key] = h
map_cluster2sequence[h].append(key)
return map_cluster2sequence, map_sequence2cluster
def ClusterPeptidesByHid(peptides):
"""cluster peptide sequences by hid."""
map_cluster2peptide = {}
map_peptide2cluster = {}
# cluster peptides by identity
# (clumsy sort, use hashes for bigger sets)
for key, sequence in peptides.items():
h = Genomics.GetHID(sequence)
if h not in map_cluster2peptide:
map_cluster2peptide[h] = []
map_peptide2cluster[key] = h
map_cluster2peptide[h].append(key)
return map_cluster2peptide, map_peptide2cluster
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 Write(map_row2col,
row_seq,
col_seq,
link,
no_gaps=False,
no_identical=False,
min_length=0,
suffix1="",
suffix2="",
outfile=None,
pair_filter=None,
format="fasta"):
"""write alignment based on map_row2col."""
status = None
filter_status = "new"
if map_row2col.getLength() == 0:
status = "empty"
if not status:
f = alignlib_lite.py_AlignmentFormatExplicit(map_row2col, row_seq,
col_seq)
row_from = map_row2col.getRowFrom()
row_to = map_row2col.getRowTo()
col_from = map_row2col.getColFrom()
col_to = map_row2col.getColTo()
row_ali, col_ali = f.mRowAlignment, f.mColAlignment
if not status:
if no_gaps:
# remove gaps from fasta
r = []
c = []
for x in range(len(row_ali)):
if row_ali[x] != "-" and col_ali[x] != "-":
r.append(row_ali[x])
c.append(col_ali[x])
row_ali = string.join(r, "")
col_ali = string.join(c, "")
if not status and len(row_ali) < min_length:
status = "length"
if not status and no_identical:
if row_ali == col_ali:
status = "identical"
if not status:
if pair_filter:
id = "%s-%s" % (link.mQueryToken, link.mSbjctToken)
if id in pair_filter:
h = Genomics.GetHID(row_ali + ";" + col_ali)
if h in pair_filter[id]:
filter_status = "old"
translation1 = Genomics.TranslateDNA2Protein(row_ali)
translation2 = Genomics.TranslateDNA2Protein(col_ali)
if "X" in translation1 or "x" in translation2:
status = "stops"
else:
status = "success"
if filter_status == "new":
if format == "fasta":
print ">%s%s %s %s\n%s\n>%s%s %s %s\n%s" % (
link.mQueryToken, suffix1, row_from, row_to, row_ali,
link.mSbjctToken, suffix2, col_from, col_to, col_ali)
elif format == "dummy":
pass
else:
raise ValueError("unknown format")
if outfile:
outfile.write("%s%s\t%s%s\t%s\t%i\t%s\n" %
(link.mQueryToken, suffix1, link.mSbjctToken, suffix2,
status, map_row2col.getLength(), filter_status))
return status
def main( argv = None ):
"""script main.
parses command line options in sys.argv, unless *argv* is given.
"""
if argv == None: argv = sys.argv
parser.add_option("-m", "--filename-map", dest="filename_map", type="string",
help="filename with mapping information.")
parser.add_option("-o", "--pattern-old", dest="pattern_old", type="string",
help="pattern for mapping new to old identifiers: extract string from old.")
parser.add_option("-n", "--pattern-new", dest="pattern_new", type="string",
help="pattern for mapping new to old identifiers: put string into new.")
parser.add_option("-g", "--genome-file", dest="genome_file", type="string",
help="genome_file.")
parser.add_option("-p", "--peptides", dest="filename_peptides", type = "string",
help="filename with peptide sequences.")
parser.add_option("-f", "--input-format", dest="input_format", type="choice",
help="format of mapping file", choices=("alignment", "offsets") )
parser.add_option("-i", "--write-missed", dest="write_missed", type="string",
help="write missed identifiers to separate file.")
parser.add_option("-a", "--filename-genes", dest="filename_genes", type="string",
help="filename with gene information.")
parser.add_option("--filename-old-peptides", dest="filename_old_peptides", type="string",
help="filename with old peptide information.")
parser.add_option("--no-renumber", dest="renumber", action="store_false",
help="do not renumber predictions.")
parser.add_option("--contig-sizes-old", dest="contig_sizes_old", type="string",
help="contig sizes for old data.")
parser.add_option("--contig-sizes-new", dest="contig_sizes_new", type="string",
help="contig sizes for new data.")
parser.add_option("--skip-errors", dest="skip_errors", action="store_true",
help="skip entries with errors.")
parser.set_defaults(
filename_map = None,
pattern_old = "(.+)",
pattern_new = "%s",
genome_file = None,
filename_peptides = None,
write_missed = None,
filename_genes = None,
filename_old_peptides = None,
renumber = True,
input_format = "alignment",
contig_sizes_old = None,
contig_sizes_new = None,
skip_errors = None
)
(options, args) = E.Start( parser, add_pipe_options = True)
predictor = PredictorExonerate()
## the different mapping criteria
map_sbjcts = {}
breakpoints = {}
################################################################################################
map_transcript2gene = {}
if options.filename_genes:
infile = open(options.filename_genes, "r")
for gene, transcript in map( lambda x: x[:-1].split("\t")[:2], filter( lambda x: x[0] != "#", infile.readlines())):
map_transcript2gene[transcript] = gene
infile.close()
################################################################################################
peptides = {}
if options.filename_peptides:
peptides = Genomics.ReadPeptideSequences( open(options.filename_peptides, "r"))
options.stdlog.write( "# read %i peptide sequences.\n" % len(peptides))
################################################################################################
## read old query sequences and compare against new query sequences
## this can be used to build a map between old and new queries
query_map_old2new = {}
if options.filename_old_peptides:
old_peptides = Genomics.ReadPeptideSequences( open(options.filename_old_peptides, "r"))
options.stdlog.write( "# read %i old peptide sequences.\n" % len(old_peptides))
query_map_old2new, unmappable, unmapped = Genomics.MapSequences( old_peptides, peptides)
options.stdlog.write( "# built map: unmappable=%i unmapped=%i.\n" % (len(unmappable), len(unmapped)))
if options.loglevel >= 2:
options.stdlog.write( "# unmappable: %s.\n" % ";".join(unmappable))
options.stdlog.write( "# unmapped: %s.\n" % ";".join(unmapped))
################################################################################################
## read old/new contig sizes for mapping positive/negative coordinates
contig_sizes_old = {}
contig_sizes_new = {}
if options.contig_sizes_old:
contig_sizes_old = Genomics.ReadContigSizes( open(options.contig_sizes_old, "r") )
if options.contig_sizes_new:
contig_sizes_new = Genomics.ReadContigSizes( open(options.contig_sizes_new, "r") )
################################################################################################
if options.filename_map:
infile = open(options.filename_map)
if options.input_format == "alignments":
for line in infile:
if line[0] == "#": continue
x, old_token, old_from, old_to, old_ali, new_from, new_to, new_ali = line[:-1].split("\t")
map_sbjcts[old_token] = (old_from, old_ali, new_from, new_ali)
if options.loglevel >= 1:
options.stdlog.write( "# read %i alignments.\n" % len(map_sbjcts))
elif options.input_format == "offsets":
## input is a list of segments and their offsets.
breakpoints, endpoints, offsets = ReadOffsets( infile )
if options.loglevel >= 1:
options.stdlog.write( "# read breakpoints for %i chromosomes.\n" % len(breakpoints))
infile.close()
################################################################################################
################################################################################################
################################################################################################
## end of input section
################################################################################################
################################################################################################
################################################################################################
rx = re.compile(options.pattern_old)
last_sbjct_token = None
ninput = 0
nerrors = 0
nerrors_map = 0
nerrors_inconsistencies = 0
nerrors_boundaries = 0
nerrors_translation = 0
nerrors_inconsequential = 0
nerrors_realigned = 0
nmapped = 0
nfiltered = 0
naligned = 0
noutput = 0
found_transcripts = {}
nduplicates = 0
output = {}
for line in sys.stdin:
if line[0] == "#": continue
entry = PredictionParser.PredictionParserEntry()
entry.Read( line )
ninput += 1
is_positive = entry.mSbjctStrand == "+"
is_error = False
## check if query token is mappable: using sequence map
if (query_map_old2new and entry.mQueryToken not in query_map_old2new):
options.stdlog.write("# skipping prediction %i: obsolete query %s\n" % (entry.mPredictionId, entry.mQueryToken) )
nfiltered += 1
continue
else:
## check if query token is mappable: using filter
if (peptides and entry.mQueryToken not in peptides):
options.stdlog.write("# skipping prediction %i: obsolete query %s\n" % (entry.mPredictionId, entry.mQueryToken) )
nfiltered += 1
continue
new_sbjct_token = options.pattern_new % rx.search(entry.mSbjctToken).groups()[0]
##########################################################################################################
## Map via alignments
if entry.mSbjctToken in map_sbjcts:
nmapped += 1
if last_sbjct_token != entry.mSbjctToken:
old_from, old_ali, new_from, new_ali = map_sbjcts[entry.mSbjctToken]
map_a2b = alignlib_lite.makeAlignmentVector()
alignlib_lite.AlignmentFormatExplicit(
int(old_from), old_ali,
int(new_from), new_ali).copy( map_a2b )
last_sbjct_token = entry.mSbjctToken
if options.loglevel >= 3:
print "#", str(entry)
print "#", map_sbjcts[entry.mSbjctToken]
sys.stdout.flush()
old_f, old_t = entry.mSbjctGenomeFrom, entry.mSbjctGenomeTo
## convert to forward coordinates:
if is_positive:
f, t= old_f, old_t
first_res, last_res = f + 1, t
else:
f, t = map_a2b.getRowTo() - old_f, map_a2b.getRowTo() - old_t
first_res, last_res = f, t + 1
## map first and last residues
mfirst_res = map_a2b.mapRowToCol( first_res )
mlast_res = map_a2b.mapRowToCol( last_res )
if (mfirst_res == 0 and old_f != 0) or (mlast_res == 0 and old_t != map_a2b.getRowTo() ):
options.stderr.write("# mapping not possible for prediction %i on %s %s:%i-%i -> %i-%i -> %i-%i -> %i-%i -> %i-%i\n" % \
(entry.mPredictionId, entry.mSbjctToken, entry.mSbjctStrand,
old_f, old_t,
f, t,
first_res, last_res,
mfirst_res, mlast_res,
f, t))
options.stderr.write("# %s\n" % str(map_sbjcts[entry.mSbjctToken]))
options.stderr.write("# %s\n" % str(entry))
options.stderr.flush()
nerrors_boundaries += 1
is_error = True
## get extended boundaries for alignment later on
while mfirst_res == 0 and first_res > 1:
first_res -= 1
mfirst_res = map_a2b.mapRowToCol(first_res)
while mlast_res == 0 and last_res < map_a2b.getRowTo():
last_res += 1
mlast_res = map_a2b.mapRowToCol(last_res)
## convert to genomic coordinates
## convert negative strand coordinates
if is_positive:
new_f = mfirst_res - 1
new_t = mlast_res
else:
new_f = mfirst_res
new_t = mlast_res - 1
new_f = map_a2b.getColTo() - new_f
new_t = map_a2b.getColTo() - new_t
## Now map the alignment.
try:
MapAlignment( entry, map_a2b )
except ValueError:
options.stderr.write("# alignment mapping not possible for prediction %i on %s %s:%i-%i -> %i-%i -> %i-%i -> %i-%i -> %i-%i -> %i-%i\n" % \
(entry.mPredictionId, entry.mSbjctToken, entry.mSbjctStrand,
old_f, old_t,
f, t,
first_res, last_res,
mfirst_res, mlast_res,
new_f, new_t,
entry.mSbjctGenomeFrom, entry.mSbjctGenomeTo))
options.stderr.write("# %s\n" % str(map_sbjcts[entry.mSbjctToken]))
options.stderr.flush()
nerrors_map += 1
is_error= True
if new_f != entry.mSbjctGenomeFrom or new_t != entry.mSbjctGenomeTo:
options.stderr.write("# mapping inconsistency for prediction %i on %s %s:%i-%i -> %i-%i -> %i-%i -> %i-%i -> %i-%i <> %i-%i\n" % \
(entry.mPredictionId, entry.mSbjctToken, entry.mSbjctStrand,
old_f, old_t,
f, t,
first_res, last_res,
mfirst_res, mlast_res,
new_f, new_t,
entry.mSbjctGenomeFrom, entry.mSbjctGenomeTo))
nerrors_inconsistencies += 1
is_error = True
##########################################################################################################
## Map via offsets
if entry.mSbjctToken in breakpoints:
old_f, old_t = entry.mSbjctGenomeFrom, entry.mSbjctGenomeTo
## convert to forward coordinates:
if is_positive:
f, t= old_f, old_t
else:
f, t = contig_sizes_old[entry.mSbjctToken] - old_t, contig_sizes_old[entry.mSbjctToken] - old_f
o1 = GetOffset( f,
breakpoints[entry.mSbjctToken],
endpoints[entry.mSbjctToken],
offsets[entry.mSbjctToken] )
o2 = GetOffset( t,
breakpoints[entry.mSbjctToken],
endpoints[entry.mSbjctToken],
offsets[entry.mSbjctToken] )
if o1 != o2:
options.stderr.write("# break within gene %s\n" % str(entry))
nerrors_map += 1
is_error = True
f += o1
t += o2
if not is_positive:
f, t = contig_sizes_new[entry.mSbjctToken] - t, contig_sizes_new[entry.mSbjctToken] - f
entry.mSbjctGenomeFrom, entry.mSbjctGenomeTo = f, t
if entry.mSbjctGenomeFrom > entry.mSbjctGenomeTo:
options.stderr.write("# mapping error: start after end %s\n" % str(entry))
nerrors_map += 1
is_error = True
##########################################################################################################
## do translation check, if genome is given
if options.genome_file:
genomic_sequence = Genomics.GetGenomicSequence( new_sbjct_token, entry.mSbjctStrand,
entry.mSbjctGenomeFrom, entry.mSbjctGenomeTo,
options.genome_file,
loglevel = 0)
map_peptide2translation, translation = Genomics.Alignment2PeptideAlignment( \
entry.mMapPeptide2Genome, entry.mQueryFrom, 0, genomic_sequence )
if re.sub("X", "", translation) != re.sub("X", "", entry.mTranslation):
options.stderr.write("# translation error for prediction %i on %s %s:%i-%i -> %i-%i <> %i-%i\n" % \
(entry.mPredictionId, entry.mSbjctToken, entry.mSbjctStrand,
old_f, old_t,
f, t,
entry.mSbjctGenomeFrom, entry.mSbjctGenomeTo))
if map_sbjcts:
options.stderr.write("# %s\n" % str(map_sbjcts[entry.mSbjctToken]))
options.stderr.write("# old=%s\n# new=%s\n" % (entry.mTranslation, translation))
options.stderr.write("# old=%s\n# new=%s\n" % (entry.mAlignmentString, Genomics.Alignment2String(entry.mMapPeptide2Genome)))
nerrors_translation += 1
is_error = True
if peptides and entry.mQueryToken in peptides:
naligned += 1
options.stdlog.write( "# aligning: %s versus %s:%s: %i-%i\n" % ( \
entry.mQueryToken,
new_sbjct_token, entry.mSbjctStrand,
entry.mSbjctGenomeFrom, entry.mSbjctGenomeTo))
# do a quick reprediction
if entry.mQueryToken in peptides:
genomic_sequence = Genomics.GetGenomicSequence( new_sbjct_token, entry.mSbjctStrand,
0, 0,
genome_file = options.genome_pattern,
loglevel = 0)
predictor.mLogLevel = 0
result = predictor(entry.mQueryToken, peptides[entry.mQueryToken],
entry.mSbjctToken, genomic_sequence,
"--exhaustive --subopt FALSE --score '%s' " % str(80),
new_f - 10, new_t + 10)
prediction_id = entry.mPredictionId
if result:
entry = result[0]
entry.mPredictionId = prediction_id
nerrors_realigned += 1
else:
if is_error:
nerrors_inconsequential += 1
entry.mSbjctToken = new_sbjct_token
## map query tokens
if query_map_old2new:
query_tokens = query_map_old2new[entry.mQueryToken]
else:
query_tokens = (entry.mQueryToken,)
if options.skip_errors and is_error:
continue
for query_token in query_tokens:
entry.mQueryToken = query_token
prediction_id = entry.mPredictionId
entry.mPredictionId = 0
hid = Genomics.GetHID( str(entry) )
if hid in output:
nduplicates += 1
continue
noutput += 1
if options.renumber: prediction_id = noutput
entry.mPredictionId = prediction_id
options.stdout.write( str(entry) + "\n")
options.stdout.flush()
found_transcripts[entry.mQueryToken] = 1
## write out found transcripts and genes
nmissed_transcripts = 0
missed_transcripts = []
found_genes = {}
if peptides:
for x in peptides.keys():
if x not in found_transcripts:
nmissed_transcripts += 1
missed_transcripts.append( x )
else:
found_genes[map_transcript2gene[x]] = 1
missed_genes = {}
nmissed_genes = 0
if map_transcript2gene:
for t in missed_transcripts:
g = map_transcript2gene[t]
if g not in found_genes:
missed_genes[g] = 1
nmissed_genes = len(missed_genes)
if options.write_missed:
outfile = open(options.write_missed, "w")
for x in missed_transcripts:
if x in unmapped:
status = "unmapped"
else:
status = "mapped"
outfile.write( "%s\t%s\t%s\n" % ("transcript", x, status ))
for x in missed_genes:
status = "unknown"
outfile.write( "%s\t%s\t%s\n" % ("gene", x, status ))
outfile.close()
options.stdlog.write("# input=%i, output=%i, filtered=%i, nduplicates=%i, mapped=%i, errors=%i\n" % (\
ninput, noutput, nfiltered, nduplicates, nmapped, nerrors ))
options.stdlog.write("# errors: inconsequental=%i, boundaries=%i, mapping=%i, inconsistencies=%i, translation=%i, realigned=%i\n" % (\
nerrors_inconsequential, nerrors_boundaries, nerrors_map, nerrors_inconsistencies, nerrors_translation, nerrors_realigned ))
options.stdlog.write("# peptides: input=%i, found=%i, missed=%i, found_genes=%i, missed_genes=%i\n" % (\
len(peptides), len(found_transcripts), nmissed_transcripts, len(found_genes), nmissed_genes) )
E.Stop()
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: optic/transcripts2links.py 2781 2009-09-10 11:33:14Z andreas $"
)
parser.add_option("-w",
"--weight",
dest="weight",
type="float",
help="weight to assign each pair.")
parser.add_option("-m",
"--method",
dest="method",
type="string",
help="method to use [genes|sequences|togenes].")
parser.add_option(
"-f",
"--filename-filter",
dest="filter",
type="string",
help="filename with filter (take only those ids in filter).")
parser.set_defaults(
method=None,
weight=0.0,
separator="|",
filter=None,
)
(options, args) = E.Start(parser)
subset = {}
if options.filter:
x = map(
lambda x: x[:-1],
filter(lambda x: x[0] != "#",
open(options.filter, "r").readlines()))
for xx in x:
subset[xx] = 1
if options.method == "genes":
transcripts = map(lambda x: x[:-1].split("\t")[0],
filter(lambda x: x[0] != "#", sys.stdin.readlines()))
if subset: transcripts = filter(lambda x: x in subset, transcripts)
transcripts = map(lambda x: x[:-1].split(options.separator),
transcripts)
transcripts.sort(lambda x, y: cmp(x[2], y[2]))
last_gene = 0
chunk = []
for species, id, gene, status in transcripts:
if last_gene != gene:
WriteLinks(sys.stdout, chunk, options.weight)
chunk = []
last_gene = gene
chunk.append(options.separator.join((species, id, gene, status)))
WriteLinks(sys.stdout, chunk, options.weight)
elif options.method == "sequences":
sequences = Genomics.ReadPeptideSequences(sys.stdin, filter=subset)
ids = sequences.keys()
hids = {}
for id, seq in sequences.items():
hid = Genomics.GetHID(seq)
if hid not in hids: hids[hid] = []
hids[hid].append(id)
for c in hids.values():
WriteLinks(sys.stdout, c, options.weight, options.separator)
elif options.method == "subsequences":
# ignore warnings from networkx/matplotlib that a display
# can not be found
with warnings.catch_warnings():
warnings.simplefilter("ignore")
import networkx
sequences = Genomics.ReadPeptideSequences(sys.stdin)
ids = sequences.keys()
## build links between all genes
graph = networkx.Graph()
graph.add_nodes_from(ids)
for x in range(len(ids) - 1):
idx = ids[x]
for y in range(x + 1, len(ids)):
idy = ids[y]
if sequences[idx] == sequences[idy] or \
sequences[idx] in sequences[idy] or \
sequences[idy] in sequences[idx]:
graph.add_edge(idx, idy)
components = networkx.connected_components(graph)
for c in components:
WriteLinks(sys.stdout, c, options.weight, options.separator)
E.Stop()
