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()
if param_loglevel >= 1:
print "# read %i cds" % (len(cds))
sys.stdout.flush()
ninput, npairs, nskipped = 0, 0, 0
for line in sys.stdin:
if line[0] == "#":
continue
if line[0] == ">":
print line[:-1]
continue
ninput += 1
link = BlastAlignments.Link()
link.Read(line)
if link.mQueryToken == link.mSbjctToken:
continue
keep = 1
if link.mQueryToken in cds and link.mSbjctToken in cds:
is_paralog, reason = IsParalogLink(link, cds[link.mQueryToken],
cds[link.mSbjctToken])
if is_paralog:
keep = 0
if param_loglevel >= 2:
print "# DISCARDED because %s: %s" % (reason, str(link))
else:
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.add_option("-n",
"--vertices",
dest="vertices",
action="append",
help="filename with vertices.")
parser.add_option("-e",
"--extra",
dest="filename_extra",
type="string",
help="filename to store extra vertices in.")
parser.add_option("-m",
"--missed",
dest="filename_missed",
type="string",
help="filename to store missed vertices in.")
parser.set_defaults(
vertices=[],
filename_extra=None,
filename_missed=None,
)
(options, args) = E.Start(parser)
if len(options.vertices) == "":
raise "please specify one set of vertices."
vertices = {}
index = 0
missed_queries = []
nvertices = [0] * len(options.vertices)
for x in range(len(options.vertices)):
f = options.vertices[x]
vv = map(lambda x: x[:-1].split("\t")[0],
filter(lambda x: x[0] != "#",
open(f, "r").readlines()))
nvertices[x] = len(vv)
missed_queries.append([])
for v in vv:
vertices[v] = [x, 0, 0]
if options.loglevel >= 1:
print "# read %i vertices from %s" % (len(vv), f)
sys.stdout.flush()
l = BlastAlignments.Link()
extra_vertices = {}
for line in sys.stdin:
if line[0] == "#":
continue
l.Read(line)
if l.mQueryToken in vertices:
vertices[l.mQueryToken][1] += 1
else:
extra_vertices[l.mQueryToken] = 1
if l.mSbjctToken in vertices:
vertices[l.mSbjctToken][2] += 1
else:
extra_vertices[l.mSbjctToken] = 1
found_queries = [0] * len(options.vertices)
found_sbjcts = [0] * len(options.vertices)
for v, vv in vertices.items():
index, nquery, nsbjct = vv
if nquery:
found_queries[index] += 1
else:
missed_queries[index].append(v)
if nsbjct:
found_sbjcts[index] += 1
headers = ("set", "name", "tvertex", "nmissed", "pmissed", "nquery",
"pquery", "nsbjct", "psbjct")
print "\t".join(headers)
for x in range(len(options.vertices)):
print "%i\t%s\t%i\t%i\t%5.2f\t%i\t%5.2f\t%i\t%5.2f" % (
x, options.vertices[x], nvertices[x], len(missed_queries[x]),
100 * float(len(missed_queries[x])) / nvertices[x],
found_queries[x], 100 * float(found_queries[x]) / nvertices[x],
found_sbjcts[x], 100 * float(found_sbjcts[x]) / nvertices[x])
print "//"
print "%i vertices not in set" % len(extra_vertices)
if options.filename_extra and len(extra_vertices) > 0:
outfile = open(options.filename_extra, "w")
for x in extra_vertices.keys():
outfile.write("%s\n" % x)
outfile.close()
if options.filename_missed:
outfile = open(options.filename_missed, "w")
for x in range(len(options.vertices)):
for y in missed_queries[x]:
outfile.write("%i\t%s\t%s\n" % (x, options.vertices[x], y))
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: graph_map_links.py 2782 2009-09-10 11:40:29Z andreas $"
)
parser.add_option("-q",
"--map-query",
dest="filename_map_query",
type="string",
help="filename with queries to map.")
parser.add_option("-s",
"--map-sbjct",
dest="filename_map_sbjct",
type="string",
help="filename with sbjcts to map.")
parser.add_option("-m",
"--multiple",
dest="multiple",
action="store_true",
help="map multiple options [%default].")
parser.add_option("-k",
"--keep-unmapped",
dest="keep_unmapped",
action="store_true",
help="keep unmapped entries [%default].")
parser.add_option("-i",
"--map-identity",
dest="map_identity",
action="store_true",
help="map by identifier [%default].")
parser.add_option(
"-n",
"--non-redundant",
dest="non_redundant",
action="store_true",
help=
"write only unique links (requires a lot of memory for large graphs) [%default]"
)
parser.set_defaults( \
filename_map_query = None,
filename_map_sbjct = None,
multiple = False,
keep_unmapped = False,
map_identity = False,
report_step = 1000000,
non_redundant = False)
(options, args) = E.Start(parser)
if options.filename_map_query:
infile = IOTools.openFile(options.filename_map_query, "r")
if options.map_identity:
map_query = readIdentityMap(infile)
else:
map_query = BlastAlignments.ReadMap(infile, options.multiple)
infile.close()
E.info('read maps for %i queries' % len(map_query))
else:
map_query = None
if options.filename_map_sbjct:
if options.filename_map_sbjct == options.filename_map_query:
map_sbjct = map_query
else:
infile = IOTools.openFile(options.filename_map_sbjct, "r")
if options.map_identity:
map_sbjct = readIdentityMap(infile)
else:
map_sbjct = BlastAlignments.ReadMap(infile, options.multiple)
infile.close()
E.info('read maps for %i sbjcts' % len(map_sbjct))
else:
map_sbjct = None
nfailed = 0
ninput = 0
nskipped = 0
noutput = 0
# number of identical/mapped links
nsame, nmapped = 0, 0
printed = {}
alignment = BlastAlignments.Map()
for line in options.stdin:
if line[0] == "#": continue
data = line[:-1].split("\t")
alignment.Read(line)
skip = False
ninput += 1
E.debug(str(map))
if options.loglevel >= 2 and ninput % options.report_step == 0:
options.stderr.write(
"# progress: ninput=%i, noutput=%i, nhash=%i\n" %
(ninput, noutput, len(printed)))
if options.multiple:
skip = False
if map_query != None:
if alignment.mQueryToken in map_query:
mq = map_query[alignment.mQueryToken]
else:
skip = True
else:
mq = [None]
if map_sbjct != None:
if alignment.mSbjctToken in map_sbjct:
ms = map_sbjct[alignment.mSbjctToken]
else:
skip = True
else:
ms = [None]
if skip:
nskipped += 1
continue
if options.map_identity:
## only if non_redundant is set, do global comparison
if not options.non_redundant: printed = {}
new_map = alignment.GetClone()
do_redundant = len(mq) > 1 or len(ms) > 1
for q in mq:
for s in ms:
new_map.mQueryToken = q
new_map.mSbjctToken = s
## check for non-redundant links for 1:many or many:many mappings
if do_redundant:
key = "%s-%i-%i-%s-%i-%i" % (
new_map.mQueryToken, new_map.mQueryFrom,
new_map.mQueryTo, new_map.mSbjctToken,
new_map.mSbjctFrom, new_map.mSbjctTo)
# hash key to save space
hkey = hashlib.md5(key).digest()
if hkey in printed: continue
printed[hkey] = 1
options.stdout.write('\t'.join([str(new_map)] +
data[9:]) + '\n')
noutput += 1
if new_map.mQueryToken == alignment.mQueryToken and \
new_map.mSbjctToken == alignment.mSbjctToken:
nsame += 1
else:
nmapped += 1
else:
for q in mq:
for s in ms:
new_map = alignment.GetClone()
E.debug(str(q))
E.debug(str(s))
is_ok = new_map.MapAlignment(q, s)
if not is_ok:
nfailed += 1
else:
options.stdout.write('\t'.join([str(new_map)] +
data[9:]) + '\n')
noutput += 1
# options.multiple == False
else:
if map_query != None:
if alignment.mQueryToken in map_query:
mq = map_query[alignment.mQueryToken]
else:
mq = None
skip = True
else:
mq = None
if map_sbjct != None:
if alignment.mSbjctToken in map_sbjct:
ms = map_sbjct[alignment.mSbjctToken]
else:
ms = None
skip = True
else:
ms = None
if skip and not options.keep_unmapped:
nskipped += 1
continue
E.debug(str(mq))
E.debug(str(ms))
if mq or ms:
is_ok = alignment.MapAlignment(mq, ms)
else:
is_ok = True
if not is_ok:
nfailed += 1
else:
options.stdout.write('\t'.join([str(alignment)] + data[9:]) +
'\n')
noutput += 1
E.info( 'ninput=%i, noutput=%i, nskipped=%i, nfailed=%i, nsame=%i, nmapped=%i' % \
(ninput, noutput, nskipped, nfailed, nsame, nmapped ))
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: 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 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: blast2fasta.py 2782 2009-09-10 11:40:29Z andreas $",
usage=globals()["__doc__"])
parser.add_option("-s",
"--sequences",
dest="filename_sequences",
type="string",
help="filename with sequences.")
parser.add_option("-f",
"--format",
dest="format",
type="string",
help="output format.")
parser.set_defaults(
filename_sequences=None,
format="fasta",
)
(options, args) = E.Start(parser)
if not options.filename_sequences:
raise "please supply filename with sequences."
sequences = Genomics.ReadPeptideSequences(
open(options.filename_sequences, "r"))
if options.loglevel >= 1:
print "# read %i sequences" % len(sequences)
for k in sequences.keys():
sequences[k] = alignlib_lite.py_makeSequence(sequences[k])
if options.loglevel >= 2:
print "# converted %i sequences" % len(sequences)
ninput, noutput, nskipped, nfailed = 0, 0, 0, 0
link = BlastAlignments.Link()
ali = alignlib_lite.py_makeAlignataVector()
for line in sys.stdin:
if line[0] == "#": continue
link.Read(line)
ninput += 1
if link.mQueryToken not in sequences or link.mSbjctToken not in sequences:
nskipped += 1
continue
ali.Clear()
alignlib_lite.py_fillAlignataCompressed(ali, link.mQueryFrom,
link.mQueryAli,
link.mSbjctFrom,
link.mSbjctAli)
result = alignlib_lite.py_writePairAlignment(
sequences[link.mQueryToken], sequences[link.mSbjctToken],
ali).split("\n")
if len(result) != 3:
nfailed += 1
if options.format == "fasta":
print ">%s %i-%i\n%s\n>%s %i-%i\n%s\n" %\
(link.mQueryToken, link.mQueryFrom, link.mQueryTo, result[0].split("\t")[1],
link.mSbjctToken, link.mSbjctFrom, link.mSbjctTo, result[1].split("\t")[1] )
noutput += 1
E.info("ninput=%i, noutput=%i, nskipped=%i, nfailed=%i" %
(ninput, noutput, nskipped, nfailed))
E.Stop()