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: optic/regions2graph.py 2754 2009-09-04 16:50:22Z andreas $",
usage=globals()["__doc__"])
parser.add_option("-b",
"--benchmark",
dest="filename_benchmark",
type="string",
help="")
parser.add_option("-y",
"--benchmark-synonyms",
dest="benchmark_synonyms",
type="string",
help="")
parser.add_option("-p",
"--peptides",
dest="filename_peptides",
type="string",
help="")
parser.add_option("-c",
"--min-coverage-query",
dest="min_coverage_query",
type="float",
help="")
parser.add_option("-s",
"--min-score",
dest="min_total_score",
type="float",
help="")
parser.add_option("-i",
"--min-percent-identity",
dest="min_percent_identity",
type="float",
help="")
parser.add_option("-o",
"--max-percent-overlap",
dest="max_percent_overlap",
type="float",
help="")
parser.add_option("--overlap-min-score",
dest="overlap_min_score",
type="float",
help="")
parser.add_option("--overlap-min-coverage",
dest="overlap_min_coverage",
type="float",
help="")
parser.add_option("--overlap-min-identity",
dest="overlap_min_identity",
type="float",
help="")
parser.add_option("--overlap-max-coverage",
dest="overlap_max_coverage",
type="float",
help="")
parser.add_option("-m",
"--max-matches",
dest="max_matches",
type="int",
help="")
parser.add_option("-j",
"--join-regions",
dest="join_regions",
type="int",
help="")
parser.add_option("--join-regions-max-regions",
dest="join_regions_max_regions",
type="int",
help="")
parser.add_option("--join-regions-max-coverage",
dest="join_regions_max_coverage",
type="float",
help="")
parser.add_option("--min-length", dest="min_length", type="int", help="")
parser.add_option("--test", dest="test", type="int", help="")
parser.add_option("--filter-queries",
dest="filename_filter_queries",
type="string",
help="")
parser.add_option("--filter-regions",
dest="filter_regions",
type="string",
help="")
parser.add_option("--conserve-memory",
dest="conserve_memory",
action="store_true",
help="")
parser.add_option("--filter-suboptimal",
dest="filter_suboptimal",
action="store_true",
help="")
parser.set_defaults(
# overlap allowed for matches on genomic region
max_percent_overlap=20,
gop=-10.0,
gep=-2.0,
# thresholds for joining regions
overlap_min_score=80,
overlap_min_coverage=80,
overlap_max_coverage=90,
overlap_min_identity=50,
# threshold for filtering bad predictions:
# minimum score
min_total_score=80,
# joining regions
join_regions=0,
# maximum coverage of query for predictions to be joined
# (This is to ensure not to join duplications. A range check
# would be better, but runs into trouble with repeats).
join_regions_max_coverage=90,
# minimum coverage of query
min_coverage_query=10,
# conserve memory
conserve_memory=0,
# minimum percent identity
min_percent_identity=0,
# minimum length
min_length=0,
max_matches=0,
filename_peptides=None,
filename_filter_queries=None,
# turn on/off various filters
filter_suboptimal=False,
filter_regions=False,
# parameters for filter of suboptimal predictions
min_relative_coverage=0.5,
min_relative_score=0.5,
min_relative_percent_identity=0.5,
# minimum difference between non-correlated conflicts to keep them
# both.
conflicts_min_difference=0.1,
# benchmarking data
benchmarks=None,
benchmark_synonyms=None,
filename_benchmark=None,
filename_benchmark_synonyms=None,
test=None,
max_intron=50000)
(options, args) = E.Start(parser, add_pipe_options=True)
if len(args) > 0:
print USAGE, "no arguments required."
sys.exit(2)
##########################################################################
# read filtering
filter_queries = {}
if options.filename_filter_queries:
for line in open(options.filename_filter_queries, "r"):
if line[0] == "#":
continue
query_token = line[:-1].split("\t")[0]
filter_queries[query_token] = True
if options.loglevel >= 1:
options.stdlog.write("# filtering for %i queries.\n" %
len(filter_queries))
##########################################################################
# read benchmarking regions
if options.filename_benchmark:
options.benchmarks = ReadBenchmarkingRegions(
open(options.filename_benchmark, "r"))
if options.loglevel >= 1:
options.stdlog.write(
"# read benchmarking regions for %i tokens\n" %
len(options.benchmarks))
sys.stdout.flush()
if options.filename_benchmark_synonyms:
infile = open(options.filename_benchmark_synonyms, "r")
options.benchmark_synonyms = {}
for line in infile:
if line[0] == "#":
continue
value, key = line[:-1].split("\t")
options.benchmark_synonyms[key] = value
else:
options.benchmark_synonyms = {}
else:
options.benchmarks = {}
options.benchmark_synonyms = {}
##########################################################################
# read peptide sequences
if options.filename_peptides:
peptide_sequences = Genomics.ReadPeptideSequences(
open(options.filename_peptides, "r"))
else:
peptide_sequences = {}
if options.conserve_memory:
old_predictions, filename_old_predictions = tempfile.mkstemp()
os.close(old_predictions)
old_predictions = PredictionFile.PredictionFile()
old_predictions.open(filename_old_predictions, "w")
else:
# array with final predictions
old_predictions = []
if options.loglevel >= 1:
options.stdlog.write("# reading predictions.\n")
sys.stdout.flush()
nread = 0
ninput = 0
for line in sys.stdin:
if line[0] == "#":
continue
entry = PredictionParser.PredictionParserEntry(expand=0)
entry.Read(line)
nread += 1
# set prediction id
if not entry.mPredictionId:
entry.mPredictionId = nread
# filter bad predictions right here in order to save memory:
if entry.score < options.min_total_score:
if options.loglevel >= 3:
options.stdlog.write(
"# PRUNING: reason: score below minimum: removing: %s\n" %
str(entry))
continue
elif entry.mQueryCoverage < options.min_coverage_query:
if options.loglevel >= 3:
options.stdlog.write(
"# PRUNING: reason: coverage below minimum: removing: %s\n"
% str(entry))
continue
elif entry.mPercentIdentity < options.min_percent_identity:
if options.loglevel >= 3:
options.stdlog.write(
"# PRUNING: reason: percent identity below minimum: removing: %s\n"
% str(entry))
continue
elif entry.mSbjctTo - entry.mSbjctFrom < options.min_length:
if options.loglevel >= 3:
options.stdlog.write(
"# PRUNING: reason: length of transcript below minimum: removing: %s\n"
% str(entry))
continue
ninput += 1
if options.test and ninput > options.test:
break
old_predictions.append(entry)
if options.loglevel >= 1:
options.stdlog.write("# predictions after input: %i\n" % ninput)
sys.stdout.flush()
if options.loglevel >= 10:
options.stdlog.write(
"############## start: predictions after input ###################################\n"
)
for x in old_predictions:
options.stdlog.write("# %s\n" % str(x))
options.stdlog.write(
"############## end: predictions after input #####################################\n"
)
sys.stdout.flush()
if ninput == 0:
options.stdlog.write("# ERROR: no predictions\n")
sys.exit(1)
##########################################################################
# set up stacks of regions
if options.conserve_memory:
old_predictions.close()
old_predictions.open(mode="r")
removed_predictions, filename_removed_predictions = tempfile.mkstemp()
os.close(removed_predictions)
removed_predictions = PredictionFile.PredictionFile()
removed_predictions.open(filename_removed_predictions, "w")
new_predictions, filename_new_predictions = tempfile.mkstemp()
os.close(new_predictions)
new_predictions = PredictionFile.PredictionFile()
new_predictions.open(filename_new_predictions, "w")
else:
removed_predictions = []
new_predictions = []
if options.benchmarks:
EvaluateBenchmark(old_predictions)
##########################################################################
# join regions
if options.join_regions and options.join_regions_max_coverage:
if options.loglevel >= 1:
options.stdlog.write(
"# joining regions: maximum distance between segments = %i and maximum query coverage = %i\n"
% (options.join_regions, options.join_regions_max_coverage))
sys.stdout.flush()
njoined = JoinRegions(old_predictions, new_predictions)
if options.conserve_memory:
ExchangeStreams(old_predictions, new_predictions)
else:
old_predictions = new_predictions
new_predictions = []
if options.loglevel >= 1:
options.stdlog.write("# predictions after joining: %i\n" % njoined)
sys.stdout.flush()
if options.loglevel >= 10:
options.stdlog.write(
"############## start: predictions after joining ###################################\n"
)
for x in old_predictions:
options.stdlog.write("# %s" % str(x))
options.stdlog.write(
"############## end: predictions after joining #####################################\n"
)
sys.stdout.flush()
else:
if options.loglevel >= 1:
options.stdlog.write("# joining regions: skipped\n")
sys.stdout.flush()
njoined = ninput
##########################################################################
# build map of best predictions
if options.filter_suboptimal:
if options.loglevel >= 1:
options.stdlog.write("# calculating best predictions\n")
sys.stdout.flush()
best_predictions = GetBestPredictions(old_predictions)
else:
best_predictions = {}
if options.loglevel >= 1:
options.stdlog.write("# calculated best predictions: %i\n" %
len(best_predictions))
sys.stdout.flush()
##########################################################################
# get regions to eliminate
filter_regions = {}
if options.filter_regions:
entry = PredictionParser.PredictionParserEntry(expand=0)
filenames = options.filter_regions.split(",")
for filename in filenames:
if options.loglevel >= 1:
options.stdlog.write("# reading regions to filter from %s.\n" %
(filename))
sys.stdout.flush()
if filename.endswith(".gz"):
infile = gzip.open(filename, "r")
else:
infile = open(filename, "r")
for line in infile:
if line[0] == "#":
continue
entry.Read(line)
exons = Exons.Alignment2Exons(
Genomics.String2Alignment(entry.mAlignmentString),
entry.mQueryFrom, entry.mSbjctGenomeFrom)
key = "%s-%s" % (entry.mSbjctToken, entry.mSbjctStrand)
if key not in filter_regions:
filter_regions[key] = []
for exon in exons:
filter_regions[key].append(
(exon.mGenomeFrom, exon.mGenomeTo))
infile.close()
for k in filter_regions.keys():
filter_regions[k].sort()
##########################################################################
# bipartite graph construction
##########################################################################
# sort predictions by genomic region
if options.conserve_memory:
old_predictions.sort(('mSbjctToken', 'mSbjctStrand',
'mSbjctGenomeFrom', 'mSbjctGenomeTo'))
else:
old_predictions.sort(lambda x, y: cmp(
(x.mSbjctToken, x.mSbjctStrand, x.mSbjctGenomeFrom, x.
mSbjctGenomeTo), (y.mSbjctToken, y.mSbjctStrand, y.
mSbjctGenomeFrom, y.mSbjctGenomeTo)))
##########################################################################
# filter predictions and resolve conflicts based on genomic overlap
# deleted segments are put in a temporary storage space.
min_from, max_from = None, None
min_to, max_to = None, None
region_id = 0
noverlaps = 0
last_prediction = None
predictions = []
region = Region()
nclusters = 0
neliminated_suboptimal = 0
neliminated_overlap = 0
noutput, nfiltered = 0, 0
for this_prediction in old_predictions:
# Filter 1: skip suboptimal predictions
if this_prediction.mQueryToken in best_predictions:
best_prediction = best_predictions[this_prediction.mQueryToken]
neliminated_suboptimal += 1
if float(
this_prediction.mQueryCoverage
) / best_prediction.mQueryCoverage < options.min_relative_coverage:
if options.loglevel >= 2:
options.stdlog.write(
"# PRUNING: reason: coverage below best: removing %s\n"
% str(this_prediction))
continue
if float(this_prediction.score
) / best_prediction.score < options.min_relative_score:
if options.loglevel >= 2:
options.stdlog.write(
"# PRUNING: reason: score below best: removing %s\n" %
str(this_prediction))
continue
if float(
this_prediction.mPercentIdentity
) / best_prediction.mPercentIdentity < options.min_relative_percent_identity:
if options.loglevel >= 2:
options.stdlog.write(
"# PRUNING: reason: percent identity below best: removing %s\n"
% str(this_prediction))
continue
neliminated_suboptimal -= 1
# Filter 2: remove predictions overlapping with certain segments
key = "%s-%s" % (this_prediction.mSbjctToken,
this_prediction.mSbjctStrand)
if key in filter_regions:
exons = Exons.Alignment2Exons(
Genomics.String2Alignment(this_prediction.mAlignmentString),
this_prediction.mQueryFrom, this_prediction.mSbjctGenomeFrom)
if CheckOverlap(map(lambda x: (x.mGenomeFrom, x.mGenomeTo), exons),
filter_regions[key]):
if options.loglevel >= 2:
options.stdlog.write(
"# PRUNING: reason: overlapping with taboo region: removing %s\n"
% str(this_prediction))
neliminated_overlap += 1
continue
try:
this_query_peptide, this_query_status, this_query_gene, this_query_transcript = \
re.split("\s+", this_prediction.mQueryToken)
except ValueError:
this_query_gene = None
# process first entry
if min_from is None:
min_from = this_prediction.mSbjctGenomeFrom
max_from = this_prediction.mSbjctGenomeFrom
max_to = this_prediction.mSbjctGenomeTo
min_to = this_prediction.mSbjctGenomeTo
predictions.append(this_prediction)
last_prediction = this_prediction
continue
overlap = min_to > this_prediction.mSbjctGenomeFrom and \
last_prediction.mSbjctToken == this_prediction.mSbjctToken and \
last_prediction.mSbjctStrand == this_prediction.mSbjctStrand
if options.loglevel >= 4:
options.stdlog.write("# from=%i, to=%i, working on: %s\n" %
(min_from, max_to, str(this_prediction)))
options.stdlog.flush()
# resolve overlap between different genes
if overlap:
noverlaps += 1
else:
region.mSbjctToken = last_prediction.mSbjctToken
region.mSbjctStrand = last_prediction.mSbjctStrand
region.mSbjctGenomeFrom = min_from
region.mSbjctGenomeTo = max_to
region_id, nxoutput, nxfiltered = ProcessRegion(
predictions, region_id, region, peptide_sequences,
filter_queries)
noutput += nxoutput
nfiltered += nxfiltered
nclusters += 1
predictions = []
min_from = this_prediction.mSbjctGenomeFrom
max_from = this_prediction.mSbjctGenomeFrom
min_to = this_prediction.mSbjctGenomeTo
max_to = this_prediction.mSbjctGenomeTo
predictions.append(this_prediction)
min_from = min(min_from, this_prediction.mSbjctGenomeFrom)
max_from = max(max_from, this_prediction.mSbjctGenomeFrom)
min_to = min(min_to, this_prediction.mSbjctGenomeTo)
max_to = max(max_to, this_prediction.mSbjctGenomeTo)
last_prediction = this_prediction
if last_prediction:
region.mSbjctToken = last_prediction.mSbjctToken
region.mSbjctStrand = last_prediction.mSbjctStrand
region.mSbjctGenomeFrom = min_from
region.mSbjctGenomeTo = max_to
region_id, nxoutput, nxfiltered = ProcessRegion(
predictions, region_id, region, peptide_sequences, filter_queries)
noutput += nxoutput
nfiltered += nxfiltered
nclusters += 1
if options.conserve_memory:
os.remove(filename_old_predictions)
os.remove(filename_new_predictions)
os.remove(filename_removed_predictions)
if options.loglevel >= 1:
options.stdlog.write(
"# pairs: nread=%i, input=%i, joined=%i, clusters=%i, regions=%i, eliminated_subopt=%i, eliminated_overlap=%i, noutput=%i, nfiltered=%i\n"
%
(nread, ninput, njoined, nclusters, region_id,
neliminated_suboptimal, neliminated_overlap, noutput, nfiltered))
E.Stop()
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: gpipe/gff2predictions.py 2021 2008-07-10 16:00:48Z andreas $",
usage=globals()["__doc__"])
parser.add_option("-t",
"--trans",
dest="trans",
help="input is translated DNA.",
action="store_true")
parser.add_option("-f",
"--format",
dest="format",
help="input format.",
type="choice",
choices=("exons", "psl", "gff"))
parser.add_option("-o",
"--output-format",
dest="output_format",
help="output format",
type="choice",
choices=('exontable', 'exons', 'predictions', 'cds',
'fasta'))
parser.add_option("-g",
"--genome-file",
dest="genome_file",
type="string",
help="filename with genomic data (indexed).")
parser.add_option(
"--predictions-file",
dest="predictions_file",
type="string",
help=
"filename with predictions. Use gene structures from this file if available."
)
parser.add_option("-i",
"--gff-field-id",
dest="gff_field_id",
type="string",
help="field for the feature id in the gff info section.")
parser.add_option(
"-p",
"--filename-peptides",
dest="filename_peptides",
type="string",
help=
"Filename with peptide sequences. If given, it is used to check the predicted translated sequences."
)
parser.add_option(
"--no-realignment",
dest="do_realignment",
action="store_false",
help="do not re-align entries that do not parse correctly.")
parser.add_option(
"--remove-unaligned",
dest="remove_unaligned",
action="store_true",
help="remove entries that have not been aligned correctly.")
parser.add_option(
"--input-coordinates",
dest="input_coordinates",
type="string",
help=
"specify input format for input coordinates [forward|both-zero|one-closed|open]."
)
parser.set_defaults(trans=False,
output_format="predictions",
format="psl",
gff_field_id='id',
input_coordinates="both-zero-open",
filename_peptides=None,
genome_file=None,
do_realignment=True,
predictions_file=None,
remove_unaligned=False)
(options, args) = E.Start(parser)
if not options.genome_file:
raise "please specify a genome file."
fasta = IndexedFasta.IndexedFasta(options.genome_file)
contig_sizes = fasta.getContigSizes()
ninput, noutput, nskipped = 0, 0, 0
nfound, nnotfound, nidentical, nmismatch, naligned, nunaligned = 0, 0, 0, 0, 0, 0
if options.filename_peptides:
peptide_sequences = Genomics.ReadPeptideSequences(
IOTools.openFile(options.filename_peptides, "r"))
predictor = Predictor.PredictorExonerate()
predictor.mLogLevel = 0
else:
peptide_sequences = None
predictor = None
converter = IndexedFasta.getConverter(options.input_coordinates)
predictions = {}
if options.predictions_file:
parser = PredictionParser.iterator_predictions(
IOTools.openFile(options.predictions_file, "r"))
for p in parser:
predictions[p.mPredictionId] = p
if options.output_format == "predictions":
if options.format == "psl":
if options.trans:
parser = PredictionParser.PredictionParserBlatTrans()
else:
parser = PredictionParser.PredictionParserBlatCDNA()
nmatches = 1
for line in sys.stdin:
if line[0] == "#":
continue
if not re.match("^[0-9]", line):
continue
try:
entries = parser.Parse((line, ))
except PredictionParser.AlignmentError, e:
print "# %s" % str(e)
print "#", line[:-1]
sys.exit(1)
for entry in entries:
entry.mPredictionId = nmatches
nmatches += 1
print str(entries)
elif options.format == "exons":
parser = PredictionParser.PredictionParserExons(
contig_sizes=contig_sizes)
else:
raise "unknown format %s for output option %s" % (
options.format, options.output_format)
if options.loglevel >= 2:
options.stdlog.write("# parsing.\n")
options.stdlog.flush()
results = parser.Parse(sys.stdin.readlines())
if options.loglevel >= 2:
options.stdlog.write("# parsing finished.\n")
options.stdlog.flush()
if options.loglevel >= 1:
options.stdlog.write(
"# parsing: ninput=%i, noutput=%i, nerrors=%i\n" %
(parser.GetNumInput(), parser.GetNumOutput(),
parser.GetNumErrors()))
for error, msg in parser.mErrors:
options.stdlog.write("# %s : %s\n" % (str(error), msg))
options.stdlog.flush()
# if genomes are given: build translation
if options.genome_file:
results.Sort(lambda x, y: cmp(x.mSbjctToken, y.mSbjctToken))
new_results = PredictionParser.Predictions()
for entry in results:
ninput += 1
if options.loglevel >= 2:
options.stdlog.write(
"# processing entry %s:%s on %s:%s %i/%i.\n" %
(entry.mPredictionId, entry.mQueryToken,
entry.mSbjctToken, entry.mSbjctStrand, ninput,
len(results)))
options.stdlog.flush()
try:
lgenome = fasta.getLength(entry.mSbjctToken)
# added 3 residues - was a problem at split codons just before the stop.
# See for example the chicken sequence ENSGALP00000002741
genomic_sequence = fasta.getSequence(
entry.mSbjctToken, entry.mSbjctStrand,
entry.mSbjctGenomeFrom,
min(entry.mSbjctGenomeTo + 3, lgenome))
except KeyError:
if options.loglevel >= 1:
options.stdlog.write(
"# did not find entry for %s on %s.\n" %
(entry.mPredictionId, entry.mSbjctToken))
nskipped += 1
continue
if predictions and entry.mPredictionId in predictions:
if options.loglevel >= 2:
options.stdlog.write(
"# substituting entry %s on %s:%s.\n" %
(entry.mPredictionId, entry.mSbjctToken,
entry.mSbjctStrand))
options.stdlog.flush()
entry = predictions[entry.mPredictionId]
exons = Exons.Alignment2Exons(entry.mMapPeptide2Genome, 0,
entry.mSbjctGenomeFrom)
entry.mMapPeptide2Translation, entry.mTranslation = Genomics.Alignment2PeptideAlignment(
Genomics.String2Alignment(entry.mAlignmentString),
entry.mQueryFrom, 0, genomic_sequence)
entry.score = entry.mMapPeptide2Translation.getColTo(
) - entry.mMapPeptide2Translation.getColFrom() + 1
(entry.mNIntrons, entry.mNFrameShifts, entry.mNGaps, entry.mNSplits, entry.mNStopCodons, entry.mNDisruptions ) = \
Genomics.CountGeneFeatures(0,
entry.mMapPeptide2Genome,
genomic_sequence)
if peptide_sequences:
if str(entry.mPredictionId) in peptide_sequences:
reference = peptide_sequences[str(
entry.mPredictionId)].upper()
translation = entry.mTranslation
nfound += 1
is_identical, nmismatches = checkIdentity(
reference, translation, options)
if is_identical:
nidentical += 1
else:
nmismatch += 1
if options.do_realignment:
if options.loglevel >= 2:
options.stdlog.write(
"# %s: mismatches..realigning in region %i:%i\n"
% (entry.mPredictionId,
entry.mSbjctGenomeFrom,
entry.mSbjctGenomeTo))
options.stdlog.flush()
result = predictor(
entry.mPredictionId, reference,
entry.mSbjctToken, genomic_sequence,
"--subopt FALSE --score '%s'" %
str(80))
# "--exhaustive --subopt FALSE --score '%s'" % str(80) )
if result:
translation = result[0].mTranslation
is_identical, nmismatches = checkIdentity(
reference, translation, options)
else:
if options.loglevel >= 2:
options.stdlog.write(
"# %s: realignment returned empty result\n"
% (entry.mPredictionId))
options.stdlog.flush()
is_identical = False
if is_identical:
naligned += 1
prediction_id = entry.mPredictionId
sbjct_genome_from = entry.mSbjctGenomeFrom
entry = result[0]
entry.mPredictionId = prediction_id
entry.mSbjctGenomeFrom += sbjct_genome_from
else:
nunaligned += 1
if options.loglevel >= 1:
options.stdlog.write(
"# %s: mismatch on %s:%s:%i-%i after realignment\n# reference =%s\n# translated=%s\n# realigned =%s\n"
%
(entry.mPredictionId,
entry.mSbjctToken,
entry.mSbjctStrand,
entry.mSbjctGenomeFrom,
entry.mSbjctGenomeTo,
reference, entry.mTranslation,
translation))
options.stdlog.flush()
if options.remove_unaligned:
nskipped += 1
continue
else:
if options.loglevel >= 2:
options.stdlog.write(
"# %s: mismatches on %s ... no realignment\n"
% (
entry.mPredictionId,
entry.mSbjctToken,
))
if options.loglevel >= 3:
options.stdlog.write(
"# %s: mismatch before realignment\n# reference =%s\n# translated=%s\n"
% (entry.mPredictionId, reference,
translation))
options.stdlog.flush()
if options.remove_unaligned:
nskipped += 1
continue
else:
nnotfound += 1
new_results.append(entry)
noutput += 1
results = new_results
if results:
options.stdout.write(str(results) + "\n")
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: gpipe/predictions2introns.py 2781 2009-09-10 11:33:14Z andreas $",
usage=globals()["__doc__"])
parser.add_option("-g",
"--genome-file",
dest="genome_file",
type="string",
help="filename with genome.")
parser.add_option("-o",
"--output-filename-summary",
dest="output_filename_summary",
type="string",
help="filename with summary information.")
parser.add_option("--skip-header",
dest="skip_header",
action="store_true",
help="skip header.")
parser.add_option(
"--fill-introns",
dest="fill_introns",
type="int",
help=
"fill intron if divisible by three and no stop codon up to a maximum length of #."
)
parser.add_option(
"--introns-max-stops",
dest="introns_max_stops",
type="int",
help="maximum number of stop codons to tolerate within an intron.")
parser.add_option("--output-format",
dest="output_format",
type="choice",
choices=("predictions", "extensions", "filled-introns"),
help="output format.")
parser.set_defaults(
genome_file="genome",
start_codons=("ATG"),
stop_codons=("TAG", "TAA", "TGA"),
skip_header=False,
)
(options, args) = E.Start(parser, add_pipe_options=True)
if len(args) > 0:
print USAGE, "no arguments required."
sys.exit(2)
fasta = IndexedFasta.IndexedFasta(options.genome_file)
p = PredictionParser.PredictionParserEntry()
ninput, noutput = 0, 0
nfilled = 0
nseqs_filled = 0
nseqs_extended = 0
left_extensions = []
right_extensions = []
filled_introns = []
if not options.skip_header:
options.stdout.write("\t".join((
"prediction_id",
"intron",
"contig",
"strand",
"start",
"end",
"length",
"nstops",
"type",
"prime5",
"prime3",
)) + "\n")
for line in sys.stdin:
if line[0] == "#": continue
ninput += 1
p.Read(line)
lsequence = fasta.getLength(p.mSbjctToken)
genomic_sequence = fasta.getSequence(p.mSbjctToken, p.mSbjctStrand,
p.mSbjctGenomeFrom,
p.mSbjctGenomeTo).upper()
exons = Exons.Alignment2Exons(p.mMapPeptide2Genome,
query_from=0,
sbjct_from=0)
new_exons = []
last_e = exons[0]
nintron = 0
for e in exons[1:]:
nintron += 1
lintron = e.mGenomeFrom - last_e.mGenomeTo
intron_is_l3 = lintron % 3 != 0
if intron_is_l3:
## get sequence, include also residues from split codons
## when checking for stop codons.
## note that e.mAlignment can sometimes be empty. This might
## be an exonerate bug. In the alignment string there are two
## consecutive exons.
if e.mAlignment and last_e.mAlignment and e.mAlignment[0][
0] == "S":
offset_left = last_e.mAlignment[-1][2]
offset_right = e.mAlignment[0][2]
else:
offset_left, offset_right = 0, 0
sequence = genomic_sequence[last_e.mGenomeTo -
offset_left:e.mGenomeFrom +
offset_right]
intron_nstops = 0
for codon in [
sequence[x:x + 3] for x in range(0, len(sequence), 3)
]:
if codon in options.stop_codons:
intron_nstops += 1
else:
intron_nstops = 0
## check for splice signals
sequence = genomic_sequence[last_e.mGenomeTo:e.mGenomeFrom]
intron_type, prime5, prime3 = Genomics.GetIntronType(sequence)
if options.loglevel >= 2:
options.stdlog.write( "\t".join(map(str, (p.mPredictionId,
nintron,
lintron,
intron_nstops,
intron_type,
genomic_sequence[last_e.mGenomeTo-6:last_e.mGenomeTo].lower() + "|" + sequence[:5] + "..." +\
sequence[-5:] + "|" + genomic_sequence[e.mGenomeFrom:e.mGenomeFrom+6].lower()) ) ) + "\n" )
options.stdout.write("\t".join(
map(str, (p.mPredictionId, nintron, p.mSbjctToken,
p.mSbjctStrand,
last_e.mGenomeTo + p.mSbjctGenomeFrom,
e.mGenomeFrom + p.mSbjctGenomeFrom, lintron,
intron_nstops, intron_type, prime5, prime3))) + "\n")
last_e = e
noutput += 1
if options.loglevel >= 1:
options.stdlog.write("# ninput=%i, noutput=%i.\n" % (\
ninput, noutput))
E.Stop()
e.mGenomeFrom, e.mGenomeTo)))
elif options.output_format == "exons":
if options.format == "exons":
parser = PredictionParser.PredictionParserExons(
contig_sizes=contig_sizes)
else:
raise "unknown format %s." % options.format
results = parser.Parse(sys.stdin.readlines())
id = 0
for entry in results:
exons = Exons.Alignment2Exons(
entry.mMapPeptide2Genome,
entry.mQueryFrom,
entry.mSbjctGenomeFrom,
)
for e in exons:
id += 1
print "\t".join(
map(str,
(entry.mQueryToken, entry.mSbjctToken,
entry.mSbjctStrand, e.frame, e.mRank, e.mPeptideFrom,
e.mPeptideTo, e.mGenomeFrom, e.mGenomeTo)))
elif options.output_format == "cds":
if options.format == "exons":
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: gpipe/predictions2cds.py 1858 2008-05-13 15:07:05Z andreas $",
usage=globals()["__doc__"])
parser.add_option("-g",
"--genome-file",
dest="genome_file",
type="string",
help="filename with genome.")
parser.add_option("-o",
"--forward-coordinates",
dest="forward_coordinates",
action="store_true",
help="input uses forward coordinates.")
parser.add_option("-f",
"--format",
dest="format",
type="choice",
choices=("default", "cds", "cdnas", "map", "gff",
"intron-fasta", "exons"),
help="output format.")
parser.add_option("-r",
"--reset-to-start",
dest="reset_to_start",
action="store_true",
help="move genomic coordinates to begin from 0.")
parser.add_option("--reset-query",
dest="reset_query",
action="store_true",
help="move peptide coordinates to begin from 0.")
parser.set_defaults(genome_file=None,
forward_coordinates=False,
format="default",
reset_to_start=False,
reset_query=False)
(options, args) = E.Start(parser, add_pipe_options=True)
if len(args) > 0:
print USAGE, "no arguments required."
sys.exit(2)
cds_id = 1
entry = PredictionParser.PredictionParserEntry()
fasta = IndexedFasta.IndexedFasta(options.genome_file)
ninput, noutput, nskipped, nerrors = 0, 0, 0, 0
for line in sys.stdin:
if line[0] == "#":
continue
if line.startswith("id"):
continue
ninput += 1
try:
entry.Read(line)
except ValueError, msg:
options.stdlog.write("# parsing failed with msg %s in line %s" %
(msg, line))
nerrors += 1
continue
cds = Exons.Alignment2Exons(entry.mMapPeptide2Genome,
query_from=entry.mQueryFrom,
sbjct_from=entry.mSbjctGenomeFrom,
add_stop_codon=0)
for cd in cds:
cd.mSbjctToken = entry.mSbjctToken
cd.mSbjctStrand = entry.mSbjctStrand
if cds[-1].mGenomeTo != entry.mSbjctGenomeTo:
options.stdlog.write(
"# WARNING: discrepancy in exon calculation!!!\n")
for cd in cds:
options.stdlog.write("# %s\n" % str(cd))
options.stdlog.write("# %s\n" % entry)
lsequence = fasta.getLength(entry.mSbjctToken)
genomic_sequence = fasta.getSequence(entry.mSbjctToken,
entry.mSbjctStrand,
entry.mSbjctGenomeFrom,
entry.mSbjctGenomeTo)
# deal with forward coordinates: convert them to negative strand
# coordinates
if options.forward_coordinates and \
entry.mSbjctStrand == "-":
entry.mSbjctGenomeFrom, entry.mSbjctGenomeTo = lsequence - \
entry.mSbjctGenomeTo, lsequence - entry.mSbjctGenomeFrom
for cd in cds:
cd.InvertGenomicCoordinates(lsequence)
# attach sequence to cds
for cd in cds:
start = cd.mGenomeFrom - entry.mSbjctGenomeFrom
end = cd.mGenomeTo - entry.mSbjctGenomeFrom
cd.mSequence = genomic_sequence[start:end]
# reset coordinates for query
if options.reset_to_start:
offset = entry.mPeptideFrom
for cd in cds:
cd.mPeptideFrom -= offset
cd.mPeptideTo -= offset
# play with coordinates
if options.reset_to_start:
offset = entry.mSbjctGenomeFrom
for cd in cds:
cd.mGenomeFrom -= offset
cd.mGenomeTo -= offset
else:
offset = 0
if options.format == "cds":
rank = 0
for cd in cds:
rank += 1
cd.mQueryToken = entry.mQueryToken
cd.mSbjctToken = entry.mSbjctToken
cd.mSbjctStrand = entry.mSbjctStrand
cd.mRank = rank
print str(cd)
if options.format == "exons":
rank = 0
for cd in cds:
rank += 1
options.stdout.write("\t".join(
map(str, (entry.mPredictionId, cd.mSbjctToken,
cd.mSbjctStrand, rank, cd.frame, cd.mPeptideFrom,
cd.mPeptideTo, cd.mGenomeFrom, cd.mGenomeTo))) +
"\n")
elif options.format == "cdnas":
print string.join(
map(str,
(entry.mPredictionId, entry.mQueryToken, entry.mSbjctToken,
entry.mSbjctStrand, entry.mSbjctGenomeFrom - offset,
entry.mSbjctGenomeTo - offset, genomic_sequence)), "\t")
elif options.format == "map":
map_prediction2genome = alignlib_lite.makeAlignmentSet()
for cd in cds:
alignlib_lite.addDiagonal2Alignment(
map_prediction2genome, cd.mPeptideFrom + 1, cd.mPeptideTo,
(cd.mGenomeFrom - offset) - cd.mPeptideFrom)
print string.join(
map(str, (entry.mPredictionId, entry.mSbjctToken,
entry.mSbjctStrand,
alignlib_lite.AlignmentFormatEmissions(
map_prediction2genome))), "\t")
elif options.format == "intron-fasta":
rank = 0
if len(cds) == 1:
nskipped += 1
continue
last = cds[0].mGenomeTo
for cd in cds[1:]:
rank += 1
key = "%s %i %s:%s:%i:%i" % (
entry.mPredictionId, rank, entry.mSbjctToken,
entry.mSbjctStrand, last, entry.mSbjctGenomeFrom)
sequence = genomic_sequence[last - entry.mSbjctGenomeFrom:cd.
mGenomeFrom -
entry.mSbjctGenomeFrom]
options.stdout.write(">%s\n%s\n" % (key, sequence))
last = cd.mGenomeTo
elif options.format == "gff-match":
print "%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\tTarget \"%s\" %i %i; Score %i; Introns %i; Frameshifts %i; Stops %i" % \
(entry.mSbjctToken,
"gpipe", "similarity",
entry.mSbjctGenomeFrom,
entry.mSbjctGenomeTo,
entry.mPercentIdentity,
entry.mSbjctStrand,
".",
entry.mQueryToken,
entry.mQueryFrom,
entry.mQueryTo,
entry.score,
entry.mNIntrons,
entry.mNFrameShifts,
entry.mNStopCodons)
elif options.format == "gff-exon":
rank = 0
for cd in cds:
rank += 1
print "%s\t%s\t%s\t%s\t%s\t%s\t%s\t%s\tTarget \"%s\" %i %i; Score %i; Rank %i/%i; Prediction %i" % \
(entry.mSbjctToken,
"gpipe", "similarity",
cd.mGenomeFrom,
cd.mGenomeTo,
entry.mPercentIdentity,
entry.mSbjctStrand,
".",
entry.mQueryToken,
cd.mPeptideFrom / 3 + 1,
cd.mPeptideTo / 3 + 1,
entry.score,
rank,
len(cds),
entry.mPredictionId)
else:
exon_from = 0
for cd in cds:
cd.mPeptideFrom = exon_from
exon_from += cd.mGenomeTo - cd.mGenomeFrom
cd.mPeptideTo = exon_from
print string.join(
map(str, (cds_id, entry.mPredictionId, cd.mPeptideFrom,
cd.mPeptideTo, cd.frame, cd.mGenomeFrom,
cd.mGenomeTo, cd.mSequence)), "\t")
cds_id += 1
noutput += 1
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: gpipe/predictions2transcripts.py 1841 2008-05-08 12:07:13Z andreas $",
usage = globals()["__doc__"] )
parser.add_option("-g", "--genome-file", dest="genome_file", type="string",
help="filename with genome." )
parser.add_option("-o", "--output-filename-summary", dest="output_filename_summary", type="string",
help="filename with summary information." )
parser.add_option( "--skip-header", dest="skip_header", action="store_true",
help="skip header." )
parser.add_option( "--start-codon-boundary", dest="start_codon_boundary", type="int",
help="maximum extension for start codon (make divisible by 3)." )
parser.add_option( "--stop-codon-boundary", dest="stop_codon_boundary", type="int",
help="maximum extension for stop codon (make divisible by 3)." )
parser.add_option( "--left-extension-mode", dest="left_extension_mode", type="choice",
choices=("first-start", "first-stop-backtrack"),
help="extension mode for 5' end.")
parser.add_option( "--fill-introns", dest="fill_introns", type="int",
help="fill intron if divisible by three and no stop codon up to a maximum length of #." )
parser.add_option( "--introns-max-stops", dest="introns_max_stops", type="int",
help="maximum number of stop codons to tolerate within an intron." )
parser.add_option( "--output-format", dest="output_format", type="choice",
choices=("predictions", "extensions", "filled-introns"),
help="output format." )
parser.set_defaults(
genome_file = "genome",
start_codons = ("ATG"),
stop_codons = ("TAG", "TAA", "TGA"),
start_codon_boundary = 9999,
stop_codon_boundary = 9999,
fill_introns = 0,
introns_max_stops = 0,
left_splice_signals = ("GT",),
right_splice_signals = ("AG",),
output_format="extensions",
left_extension_mode = "first-start",
skip_header = False,
output_filename_summary = None,
)
(options, args) = E.Start( parser, add_pipe_options = True )
if len(args) > 0:
print USAGE, "no arguments required."
sys.exit(2)
options.start_codon_boundary = int(options.start_codon_boundary / 3)
options.stop_codon_boundary = int(options.stop_codon_boundary / 3)
fasta = IndexedFasta.IndexedFasta( options.genome_file )
p = PredictionParser.PredictionParserEntry()
ninput, noutput = 0, 0
nfilled = 0
nseqs_filled = 0
nseqs_extended = 0
left_extensions = []
right_extensions = []
filled_introns = []
if not options.skip_header:
if options.output_format == "predictions":
options.stdout.write( Prediction.Prediction().getHeader() + "\n" )
elif options.output_format == "filled-introns":
options.stdout.write("\t".join( ("prediction_id",
"intron",
"peptide_sequence",
"genomic_sequence") ) + "\n" )
if options.output_filename_summary:
outfile_summary = open (options.output_filename_summary, "w" )
outfile_summary.write( "id\ttype\tnumber\tlength\tfrom\tto\tsequence\n" )
else:
outfile_summary = None
for line in options.stdin:
if line[0] == "#": continue
ninput += 1
p.Read(line)
lsequence = fasta.getLength( p.mSbjctToken )
genome_from = max( 0, p.mSbjctGenomeFrom - options.start_codon_boundary)
genome_to = min( lsequence, p.mSbjctGenomeTo + options.stop_codon_boundary)
genomic_sequence = fasta.getSequence( p.mSbjctToken, p.mSbjctStrand,
genome_from,
genome_to ).upper()
########################################################################
########################################################################
########################################################################
## Do extensions
if options.start_codon_boundary or options.stop_codon_boundary:
extension_start = p.mSbjctGenomeFrom - genome_from
extension_stop = genome_to - p.mSbjctGenomeTo
fragment_to = extension_start + p.mSbjctGenomeTo - p.mSbjctGenomeFrom
lfragment = len(genomic_sequence)
########################################################################
########################################################################
########################################################################
## find start codon
start = extension_start
found_start = False
if options.left_extension_mode == "first-start":
found_start, start = findCodonReverse( genomic_sequence,
start,
options.start_codons,
options.stop_codons )
elif options.left_extension_mode == "first-stop-backtrack":
if genomic_sequence[start:start+3] in options.start_codons:
found_start = True
else:
found_start, start = findCodonReverse( genomic_sequence,
start,
options.stop_codons )
if found_start:
E.info("prediction %s: stop found at %i (%i) backtracking ..." % ( p.mPredictionId, start, extension_start - start) )
## bracktrack to first start codon
found_start = False
while start < extension_start:
start += 3
if genomic_sequence[start:start+3] in options.start_codons:
found_start = True
break
else:
start = extension_start
if found_start:
E.info("start codon found at %i (%i)." % ( start, extension_start - start) )
else:
E.info("no start codon found." )
else:
E.info("prediction %s: no stop found ... backtracking to start codon." % ( p.mPredictionId ) )
found_start, start = findCodonReverse( genomic_sequence, start, options.start_codons )
E.info("prediction %s: no start codon found." % ( p.mPredictionId ) )
if found_start:
start += genome_from
else:
start = p.mSbjctGenomeFrom
dstart = p.mSbjctGenomeFrom - start
########################################################################
########################################################################
########################################################################
## find stop codon
## stop points to the beginning of the codon, thus the stop codon will
## not be part of the sequence.
stop = fragment_to
found_stop = 0
while stop < lfragment and \
genomic_sequence[stop:stop+3] not in ("NNN", "XXX"):
if genomic_sequence[stop:stop+3] in options.stop_codons:
found_stop = 1
break
stop += 3
if found_stop:
stop += genome_from
else:
stop = p.mSbjctGenomeTo
dstop = stop - p.mSbjctGenomeTo
########################################################################
########################################################################
########################################################################
## build new prediction
map_peptide2genome = []
if dstart: map_peptide2genome.append( ("G", 0, dstart) )
map_peptide2genome += p.mMapPeptide2Genome
if dstop: map_peptide2genome.append( ("G", 0, dstop) )
E.info("prediction %s: extension: found_start=%i, found_stop=%i, left=%i, right=%i" % ( p.mPredictionId, found_start, found_stop, dstart, dstop ) )
## save results
p.mMapPeptide2Genome = map_peptide2genome
p.mAlignmentString = Genomics.Alignment2String( map_peptide2genome )
p.mSbjctGenomeFrom -= dstart
p.mSbjctGenomeTo += dstop
p.mSbjctFrom += dstart / 3
p.mSbjctTo += dstart / 3 + dstop / 3
if dstart or dstop:
if dstart: left_extensions.append( dstart )
if dstop: right_extensions.append( dstop )
nseqs_extended += 1
## update genomic sequence because borders might have changed.
genomic_sequence = fasta.getSequence( p.mSbjctToken,
p.mSbjctStrand,
p.mSbjctGenomeFrom,
p.mSbjctGenomeTo ).upper()
if options.fill_introns:
has_filled = False
exons = Exons.Alignment2Exons( p.mMapPeptide2Genome,
query_from = 0,
sbjct_from = 0 )
new_exons = []
last_e = exons[0]
nintron = 0
for e in exons[1:]:
nintron += 1
lintron = e.mGenomeFrom - last_e.mGenomeTo
if lintron > options.fill_introns or (lintron) % 3 != 0:
E.debug( "prediction %s: intron %i of size %i discarded." % \
(p.mPredictionId,
nintron, lintron ) )
new_exons.append(last_e)
last_e = e
continue
## get sequence, include also residues from split codons
## when checking for stop codons.
if e.mAlignment[0][0] == "S":
offset_left = last_e.mAlignment[-1][2]
offset_right = e.mAlignment[0][2]
else:
offset_left, offset_right = 0, 0
sequence = genomic_sequence[last_e.mGenomeTo - offset_left:e.mGenomeFrom+offset_right]
## check for splice sites
for signal in options.left_splice_signals:
if sequence[offset_left:offset_left+len(signal)] == signal:
left_signal = True
break
else:
left_signal = False
for signal in options.right_splice_signals:
if sequence[-(len(signal)+offset_right):-offset_right] == signal:
right_signal = True
break
else:
right_signal = False
nstops, ngaps = 0, 0
for codon in [ sequence[x:x+3] for x in range(0,len(sequence),3) ]:
if codon in options.stop_codons: nstops += 1
if "N" in codon.upper(): ngaps += 1
E.debug( "prediction %s: intron %i of size %i (%i-%i) (%s:%s:%i:%i): stops=%i, gaps=%i, signals=%s,%s." % \
(p.mPredictionId,
nintron, lintron,
offset_left, offset_right,
p.mSbjctToken, p.mSbjctStrand,
p.mSbjctGenomeFrom + last_e.mGenomeTo,
p.mSbjctGenomeFrom + e.mGenomeFrom,
nstops,
ngaps,
left_signal, right_signal ) )
if nstops + ngaps > options.introns_max_stops:
new_exons.append(last_e)
last_e = e
continue
E.info( "prediction %s: filling intron %i of size %i: stops=%i, gaps=%i, signals=%s,%s" % \
(p.mPredictionId,
nintron, lintron,
nstops,
ngaps,
left_signal, right_signal))
e.Merge( last_e )
has_filled = True
nfilled += 1
last_e = e
if options.output_format == "filled-introns":
options.stdout.write( "\t".join( map(str, ( p.mPredictionId,
nintron,
Genomics.TranslateDNA2Protein( sequence ),
sequence ) ) ) + "\n" )
filled_introns.append(lintron)
p.mNIntrons -= 1
new_exons.append(last_e)
if has_filled: nseqs_filled += 1
Exons.UpdatePeptideCoordinates( new_exons )
p.mMapPeptide2Genome = Exons.Exons2Alignment( new_exons )
p.mAlignmentString = Genomics.Alignment2String( p.mMapPeptide2Genome )
## build translated sequence
p.mMapPeptide2Translation, p.mTranslation = Genomics.Alignment2PeptideAlignment( \
p.mMapPeptide2Genome, p.mQueryFrom, 0, genomic_sequence )
## output info
if options.output_format == "predictions":
options.stdout.write( str(p) + "\n" )
elif options.output_format == "extensions":
if found_start: found_start = 1
if found_stop: found_stop = 1
options.stdout.write( "\t".join( map(str, ( p.mPredictionId,
found_start, found_stop,
dstart, dstop,
p.mTranslation,
p.mSbjctGenomeFrom, p.mSbjctGenomeTo,
p.mAlignmentString ))) + "\n" )
noutput += 1
options.stdout.flush()
E.info("stats : %s" % "\t".join(Stats.DistributionalParameters().getHeaders() ))
E.info("left : %s" % str(Stats.DistributionalParameters(left_extensions)) )
E.info("right : %s" % str(Stats.DistributionalParameters(right_extensions)) )
E.info("introns: %s" % str(Stats.DistributionalParameters(filled_introns)) )
E.info("ninput=%i, noutput=%i, nextended=%i, nfilled=%i, nexons_filled=%i" % (\
ninput, noutput, nseqs_extended, nseqs_filled, nfilled))
E.Stop()
def ResolveExonOverlaps(gene_id, predictions):
"""resolve overlaps between predictions based
on exonic overlap."""
all_exons = []
n = 1
if len(predictions) == 0:
return gene_id
for p in predictions:
exons = Exons.Alignment2Exons(Genomics.String2Alignment(
p.mAlignmentString),
query_from=0,
sbjct_from=p.mSbjctGenomeFrom)
for exon in exons:
all_exons.append((exon.mGenomeFrom, exon.mGenomeTo, n))
else:
all_exons.append((p.mSbjctGenomeFrom, p.mSbjctGenomeTo, n))
n += 1
map_prediction2gene = range(0, len(predictions) + 1)
map_gene2predictions = [None]
for x in range(1, len(predictions) + 1):
map_gene2predictions.append([x])
all_exons.sort()
# print all_exons
# cluster exons by overlap
last_exon_from, last_exon_to, last_p = all_exons[0]
for exon_from, exon_to, p in all_exons[1:]:
# if overlap
overlap = min(exon_to, last_exon_to) - max(exon_from, last_exon_from)
if overlap and param_exon_identity:
overlap = (exon_from == last_exon_from) and (exon_to
== last_exon_to)
if overlap > 0:
# print "# overlap between %i and %i" % (p, last_p)
# rewire pointers to point to gene of previous prediction
# if they belong to different genes
new_g = map_prediction2gene[last_p]
old_g = map_prediction2gene[p]
if new_g != old_g:
for x in map_gene2predictions[old_g]:
map_gene2predictions[new_g].append(x)
map_prediction2gene[x] = new_g
map_gene2predictions[old_g] = []
# if no overlap: create new gene, if predictions has no gene
# associated with it yet.
else:
# print "# no overlap between %i and %i" % (p, last_p)
if not map_prediction2gene[p]:
map_prediction2gene[p] = len(map_gene2predictions)
map_gene2predictions.append([p])
if param_exon_identity:
last_exon_to = exon_to
last_exon_from = exon_from
else:
last_exon_to = max(last_exon_to, exon_to)
last_p = p
for x in range(1, len(map_gene2predictions)):
if map_gene2predictions[x]:
for p in map_gene2predictions[x]:
print "%i\t%i" % (gene_id, predictions[p - 1].mPredictionId)
gene_id += 1
return gene_id
alignlib_lite.ALIGNMENT_LOCAL, param_gop, param_gep)
map_reference2target = alignlib_lite.makeAlignmentVector()
assignment_id = 0
for line in cr.fetchall():
reference = PredictionParser.PredictionParserEntry()
reference.FillFromTable(line)
ct = dbhandle.cursor()
ct.execute(statement %
(param_tablename_predictions_target, reference.mSbjctToken,
reference.mSbjctStrand, reference.mSbjctGenomeFrom,
reference.mSbjctGenomeTo))
reference_exons = Exons.Alignment2Exons(reference.mMapPeptide2Genome,
0, reference.mSbjctFrom)
for line2 in ct.fetchall():
target = PredictionParser.PredictionParserEntry()
target.FillFromTable(line2)
target_exons = Exons.Alignment2Exons(target.mMapPeptide2Genome, 0,
target.mSbjctFrom)
## check for exon overlap
rr, tt = 0, 0
overlap = 0
while rr < len(reference_exons) and tt < len(target_exons):
r = reference_exons[rr]
t = target_exons[tt]