Esempio n. 1
0
def insert(args):
    """
    %prog insert candidates.bed gaps.bed chrs.fasta unplaced.fasta

    Insert scaffolds into assembly.
    """
    from jcvi.formats.agp import mask, bed
    from jcvi.formats.sizes import agp

    p = OptionParser(insert.__doc__)
    opts, args = p.parse_args(args)

    if len(args) != 4:
        sys.exit(not p.print_help())

    candidates, gapsbed, chrfasta, unplacedfasta = args
    refinedbed = refine([candidates, gapsbed])
    sizes = Sizes(unplacedfasta).mapping
    cbed = Bed(candidates)
    corder = cbed.order
    gbed = Bed(gapsbed)
    gorder = gbed.order

    gpbed = Bed()
    gappositions = {}  # (chr, start, end) => gapid

    fp = open(refinedbed)
    gap_to_scf = defaultdict(list)
    seen = set()
    for row in fp:
        atoms = row.split()
        if len(atoms) <= 6:
            continue
        unplaced = atoms[3]
        strand = atoms[5]
        gapid = atoms[9]
        if gapid not in seen:
            seen.add(gapid)
            gi, gb = gorder[gapid]
            gpbed.append(gb)
            gappositions[(gb.seqid, gb.start, gb.end)] = gapid
        gap_to_scf[gapid].append((unplaced, strand))

    gpbedfile = "candidate.gaps.bed"
    gpbed.print_to_file(gpbedfile, sorted=True)

    agpfile = agp([chrfasta])
    maskedagpfile = mask([agpfile, gpbedfile])
    maskedbedfile = maskedagpfile.rsplit(".", 1)[0] + ".bed"
    bed([maskedagpfile, "--outfile={0}".format(maskedbedfile)])

    mbed = Bed(maskedbedfile)
    finalbed = Bed()
    for b in mbed:
        sid = b.seqid
        key = (sid, b.start, b.end)
        if key not in gappositions:
            finalbed.add("{0}\n".format(b))
            continue

        gapid = gappositions[key]
        scfs = gap_to_scf[gapid]

        # For scaffolds placed in the same gap, sort according to positions
        scfs.sort(key=lambda x: corder[x[0]][1].start + corder[x[0]][1].end)
        for scf, strand in scfs:
            size = sizes[scf]
            finalbed.add("\t".join(str(x) for x in (scf, 0, size, sid, 1000, strand)))

    finalbedfile = "final.bed"
    finalbed.print_to_file(finalbedfile)

    # Clean-up
    toclean = [gpbedfile, agpfile, maskedagpfile, maskedbedfile]
    FileShredder(toclean)
Esempio n. 2
0
def insert(args):
    """
    %prog insert candidates.bed gaps.bed chrs.fasta unplaced.fasta

    Insert scaffolds into assembly.
    """
    from jcvi.formats.agp import mask, bed
    from jcvi.formats.sizes import agp

    p = OptionParser(insert.__doc__)
    opts, args = p.parse_args(args)

    if len(args) != 4:
        sys.exit(not p.print_help())

    candidates, gapsbed, chrfasta, unplacedfasta = args
    refinedbed = refine([candidates, gapsbed])
    sizes = Sizes(unplacedfasta).mapping
    cbed = Bed(candidates)
    corder = cbed.order
    gbed = Bed(gapsbed)
    gorder = gbed.order

    gpbed = Bed()
    gappositions = {}  # (chr, start, end) => gapid

    fp = open(refinedbed)
    gap_to_scf = defaultdict(list)
    seen = set()
    for row in fp:
        atoms = row.split()
        unplaced = atoms[3]
        strand = atoms[5]
        gapid = atoms[9]
        if gapid not in seen:
            seen.add(gapid)
            gi, gb = gorder[gapid]
            gpbed.append(gb)
            gappositions[(gb.seqid, gb.start, gb.end)] = gapid
        gap_to_scf[gapid].append((unplaced, strand))

    gpbedfile = "candidate.gaps.bed"
    gpbed.print_to_file(gpbedfile, sorted=True)

    agpfile = agp([chrfasta])
    maskedagpfile = mask([agpfile, gpbedfile])
    maskedbedfile = maskedagpfile.rsplit(".", 1)[0] + ".bed"
    bed([maskedagpfile, "--outfile={0}".format(maskedbedfile)])

    mbed = Bed(maskedbedfile)
    beds = []
    for b in mbed:
        sid = b.seqid
        key = (sid, b.start, b.end)
        if key not in gappositions:
            beds.append(b)
            continue

        gapid = gappositions[key]
        scfs = gap_to_scf[gapid]

        # For scaffolds placed in the same gap, sort according to positions
        scfs.sort(key=lambda x: corder[x[0]][1].start + corder[x[0]][1].end)
        for scf, strand in scfs:
            size = sizes[scf]
            beds.append(BedLine("\t".join(str(x) for x in \
                    (scf, 0, size, sid, 1000, strand))))

    finalbed = Bed()
    finalbed.extend(beds)
    finalbedfile = "final.bed"
    finalbed.print_to_file(finalbedfile)

    # Clean-up
    toclean = [gpbedfile, agpfile, maskedagpfile, maskedbedfile]
    FileShredder(toclean)