def get_transcripts(annot,
coding_only,
genes=None,
only_longest=False,
coding_first=False):
transcripts_dict = Transcript.extract_transcripts(annot)
transcripts = []
if only_longest:
by_gene = defaultdict(list)
for transcript in transcripts_dict.values():
by_gene[transcript.gene].append(transcript)
for gene in by_gene.keys():
transcripts.append(get_longest(by_gene[gene], coding_first))
else:
transcripts = transcripts_dict.values()
if genes and type(genes) is Set:
transcripts = [t for t in transcripts if t.gene in genes]
captured_genes = Set([t.gene for t in transcripts])
for gene in genes - Set([t.gene for t in transcripts]):
print "can't find %s from gtf" % gene
if coding_only:
transcripts = [t for t in transcripts if t.is_coding()]
return transcripts
def main():
args = parse_args()
bam = create_pysam_bam(args.bam)
query_fasta = create_pysam_fasta(args.query_fasta)
genome_fasta = create_pysam_fasta(args.genome_fasta)
transcripts_dict = Transcript.extract_transcripts(args.gtf)
annot_tabix = create_pysam_tabix(args.gtf)
genome_bam = None
if args.genome_bam:
genome_bam = create_pysam_bam(args.genome_bam)
em = ExonMapper(annot_tabix,
transcripts_dict,
genome_fasta,
debug=args.debug)
annots = [args.gtf]
if args.suppl_annot:
annots.extend(args.suppl_annot)
accessory_known_features = extract_features(annots)
mappings, junc_adjs, events = em.map_aligns(
bam,
query_fasta,
genome_fasta,
accessory_known_features=accessory_known_features,
max_diff=args.max_diff_splice)
juncs_merged = Adjacency.merge(junc_adjs)
events_merged = Adjacency.merge(events)
if args.r2c:
all_adjs = []
if juncs_merged:
all_adjs.extend(juncs_merged)
if events_merged:
all_adjs.extend(events_merged)
if all_adjs:
find_support(all_adjs,
args.r2c,
args.query_fasta,
num_procs=args.nproc,
debug=args.debug)
if events_merged:
filter_events(events_merged, args.min_support)
if genome_bam:
corroborate_genome(events_merged, genome_bam)
cmd = ' '.join(sys.argv)
time = datetime.datetime.now().strftime("%Y-%m-%d:%H:%M:%S")
software = '%s %s' % (pv.__name__, pv.__version__)
em.output_mappings(mappings, '%s/mappings.tsv' % args.outdir)
em.output_juncs(juncs_merged, '%s/junctions.bed' % args.outdir)
em.output_events(events_merged,
'%s/novel_splicing.bedpe' % args.outdir,
header=(software, '%s %s' % (time, cmd)))
def main():
args = parse_args()
bam = create_pysam_bam(args.bam)
query_fasta = create_pysam_fasta(args.query_fasta)
genome_fasta = create_pysam_fasta(args.genome_fasta)
transcripts_dict = Transcript.extract_transcripts(args.gtf)
annot_tabix = create_pysam_tabix(args.gtf)
genome_bam = None
if args.genome_bam:
genome_bam = create_pysam_bam(args.genome_bam)
em = ExonMapper(annot_tabix,
transcripts_dict,
genome_fasta,
debug = args.debug)
annots = [args.gtf]
if args.suppl_annot:
annots.extend(args.suppl_annot)
accessory_known_features = extract_features(annots)
mappings, junc_adjs, events = em.map_aligns(bam,
query_fasta,
genome_fasta,
accessory_known_features=accessory_known_features,
max_diff=args.max_diff_splice)
juncs_merged = Adjacency.merge(junc_adjs)
events_merged = Adjacency.merge(events)
if args.r2c:
all_adjs = []
if juncs_merged:
all_adjs.extend(juncs_merged)
if events_merged:
all_adjs.extend(events_merged)
if all_adjs:
find_support(all_adjs, args.r2c, args.query_fasta, num_procs=args.nproc, debug=args.debug)
if events_merged:
filter_events(events_merged, args.min_support)
if genome_bam:
corroborate_genome(events_merged, genome_bam)
cmd = ' '.join(sys.argv)
time = datetime.datetime.now().strftime("%Y-%m-%d:%H:%M:%S")
software = '%s %s' % (pv.__name__, pv.__version__)
em.output_mappings(mappings, '%s/mappings.tsv' % args.outdir)
em.output_juncs(juncs_merged, '%s/junctions.bed' % args.outdir)
em.output_events(events_merged, '%s/novel_splicing.bedpe' % args.outdir, header=(software, '%s %s' % (time, cmd)))
def main():
args = parse_args()
# extract unmapped read pairs
r2c_bam = pysam.AlignmentFile(args.r2c)
unmapped_fastas = find_unmapped_mates(r2c_bam, args.outdir)
# align unmapped reads
unmapped_bam_file = align_unmapped_mates(unmapped_fastas, args.transcripts_fasta, args.nthreads, args.outdir)
# extract fusions
unmapped_bam = pysam.AlignmentFile(unmapped_bam_file)
transcripts_dict = Transcript.extract_transcripts(args.gtf)
adjs = find_discordant_pairs(unmapped_bam, transcripts_dict, pysam.FastaFile(args.genome_fasta), min_pairs=args.min_pairs)
Adjacency.report_events(adjs, '{}/discordant_pairs.bedpe'.format(args.outdir))
if not args.no_cleanup:
cleanup(args.outdir)
def main():
args = parse_args()
# extract unmapped read pairs
r2c_bam = pysam.AlignmentFile(args.r2c)
unmapped_fastas = find_unmapped_mates(r2c_bam, args.outdir)
# align unmapped reads
unmapped_bam_file = align_unmapped_mates(unmapped_fastas, args.transcripts_fasta, args.nthreads, args.outdir)
# extract fusions
unmapped_bam = pysam.AlignmentFile(unmapped_bam_file)
transcripts_dict = Transcript.extract_transcripts(args.gtf)
adjs = find_discordant_pairs(unmapped_bam, transcripts_dict, pysam.FastaFile(args.genome_fasta), min_pairs=args.min_pairs)
Adjacency.report_events(adjs, '%s/discordant_pairs.bedpe' % args.outdir)
if not args.no_cleanup:
cleanup(args.outdir)
def get_transcripts(annot, coding_only, genes=None, only_longest=False, coding_first=False):
transcripts_dict = Transcript.extract_transcripts(annot)
transcripts = []
if only_longest:
by_gene = defaultdict(list)
for transcript in transcripts_dict.values():
by_gene[transcript.gene].append(transcript)
for gene in by_gene.keys():
transcripts.append(get_longest(by_gene[gene], coding_first))
else:
transcripts = transcripts_dict.values()
if genes and type(genes) is Set:
transcripts = [t for t in transcripts if t.gene in genes]
captured_genes = Set([t.gene for t in transcripts])
for gene in genes - Set([t.gene for t in transcripts]):
print "can't find %s from gtf" % gene
if coding_only:
transcripts = [t for t in transcripts if t.is_coding()]
return transcripts
def main():
args = parse_args()
gbam = create_pysam_bam(args.gbam)
tbam = create_pysam_bam(args.tbam)
query_fasta = create_pysam_fasta(args.query_fasta)
genome_fasta = create_pysam_fasta(args.genome_fasta)
transcripts_fasta = create_pysam_fasta(args.transcripts_fasta)
transcripts_dict = Transcript.extract_transcripts(args.gtf)
annot_tabix = create_pysam_tabix(args.gtf)
sf = SVFinder(genome_fasta,
annot_tabix,
transcripts_dict,
args.outdir,
probe_len=args.probe_len,
debug=args.debug)
events = {'via_genome': {}, 'via_transcripts': {}}
mappings = {'via_genome': {}, 'via_transcripts': {}}
gene_hits = None
if gbam and annot_tabix:
events['via_genome'], mappings['via_genome'] = sf.find_events(
gbam,
query_fasta,
genome_fasta,
'genome',
min_indel_size=args.min_indel_size,
min_dup_size=args.min_dup_size,
min_indel_flanking=args.min_indel_flanking,
no_utr=args.no_utr,
max_novel_len=args.max_novel_len,
max_homol_len=args.max_homol_len,
only_sense_fusion=not args.include_nonsense_fusion,
only_exon_bound_fusion=not args.include_non_exon_bound_fusion,
only_coding_fusion=not args.include_noncoding_fusion,
only_fusions=args.only_fusions)
if tbam:
events['via_transcripts'], mappings[
'via_transcripts'] = sf.find_events(
tbam,
query_fasta,
transcripts_fasta,
'transcripts',
external_mappings=mappings['via_genome'],
min_indel_size=args.min_indel_size,
min_dup_size=args.min_dup_size,
min_indel_flanking=args.min_indel_flanking,
no_utr=args.no_utr,
no_indels=True,
max_novel_len=args.max_novel_len,
max_homol_len=args.max_homol_len,
only_sense_fusion=not args.include_nonsense_fusion,
only_exon_bound_fusion=not args.include_non_exon_bound_fusion,
only_coding_fusion=not args.include_noncoding_fusion,
only_fusions=args.only_fusions)
# combine events from genome and transcriptome alignments
events_combined = combine_events(events, mappings)
# merge identical events from different contigs
events_merged = Adjacency.merge(events_combined)
# filter by checking probe and subseq alignments
if events_merged and args.genome_index and len(args.genome_index) == 2:
if not args.disable_subseq_filtering:
sf.filter_subseqs(events_merged,
query_fasta,
args.genome_index[0],
args.genome_index[1],
args.outdir,
subseq_len=args.subseq_len,
debug=args.debug)
if not args.disable_probe_filtering:
sf.filter_probes(events_merged,
args.genome_index[0],
args.genome_index[1],
args.outdir,
args.probe_len,
debug=args.debug)
# read support
if args.r2c:
find_support(events_merged,
args.r2c,
args.query_fasta,
min_overlap=args.min_overhang,
num_procs=args.nproc,
debug=args.debug)
events_filtered = [
event for event in events_merged
if event.spanning >= args.min_support
]
else:
events_filtered = events_merged
# determine if events are in- or out-of-frame
sf.set_frame(events_filtered, query_fasta, genome_fasta)
# report (with meta data)
cmd = ' '.join(sys.argv)
time = datetime.datetime.now().strftime("%Y-%m-%d:%H:%M:%S")
software = '{} {}'.format(pv.__name__, pv.__version__)
Adjacency.report_events(events_filtered,
'{}/sv.bedpe'.format(args.outdir),
sort_by_coord=args.sort_by_coord,
header=(software, '{} {}'.format(time, cmd)))
def main():
args = parse_args()
gbam = create_pysam_bam(args.gbam)
tbam = create_pysam_bam(args.tbam)
query_fasta = create_pysam_fasta(args.query_fasta)
genome_fasta = create_pysam_fasta(args.genome_fasta)
transcripts_fasta = create_pysam_fasta(args.transcripts_fasta)
transcripts_dict = Transcript.extract_transcripts(args.gtf)
annot_tabix = create_pysam_tabix(args.gtf)
sf = SVFinder(genome_fasta, annot_tabix, transcripts_dict, args.outdir, probe_len=args.probe_len, debug=args.debug)
events = {'via_genome': {}, 'via_transcripts': {}}
mappings = {'via_genome': {}, 'via_transcripts': {}}
gene_hits = None
if gbam and annot_tabix:
events['via_genome'], mappings['via_genome'] = sf.find_events(gbam,
query_fasta,
genome_fasta,
'genome',
min_indel_size=args.min_indel_size,
min_dup_size=args.min_dup_size,
min_indel_flanking=args.min_indel_flanking,
no_utr=args.no_utr,
max_novel_len=args.max_novel_len,
max_homol_len=args.max_homol_len,
only_sense_fusion=not args.include_nonsense_fusion,
only_exon_bound_fusion=not args.include_non_exon_bound_fusion,
only_coding_fusion=not args.include_noncoding_fusion,
only_fusions=args.only_fusions
)
if tbam:
events['via_transcripts'], mappings['via_transcripts'] = sf.find_events(tbam,
query_fasta,
transcripts_fasta,
'transcripts',
external_mappings=mappings['via_genome'],
min_indel_size=args.min_indel_size,
min_dup_size=args.min_dup_size,
min_indel_flanking=args.min_indel_flanking,
no_utr=args.no_utr,
no_indels=True,
max_novel_len=args.max_novel_len,
max_homol_len=args.max_homol_len,
only_sense_fusion=not args.include_nonsense_fusion,
only_exon_bound_fusion=not args.include_non_exon_bound_fusion,
only_coding_fusion=not args.include_noncoding_fusion,
only_fusions=args.only_fusions
)
# combine events from genome and transcriptome alignments
events_combined = combine_events(events, mappings)
# merge identical events from different contigs
events_merged = Adjacency.merge(events_combined)
# filter by checking probe and subseq alignments
if events_merged and args.genome_index and len(args.genome_index) == 2:
if not args.disable_subseq_filtering:
sf.filter_subseqs(events_merged, query_fasta, args.genome_index[0], args.genome_index[1], args.outdir,
subseq_len=args.subseq_len, debug=args.debug)
if not args.disable_probe_filtering:
sf.filter_probes(events_merged, args.genome_index[0], args.genome_index[1], args.outdir, args.probe_len, debug=args.debug)
# read support
if args.r2c:
find_support(events_merged, args.r2c, args.query_fasta, min_overlap=args.min_overhang, num_procs=args.nproc, debug=args.debug)
events_filtered = [event for event in events_merged if event.spanning >= args.min_support]
else:
events_filtered = events_merged
# determine if events are in- or out-of-frame
sf.set_frame(events_filtered, query_fasta, genome_fasta)
# report (with meta data)
cmd = ' '.join(sys.argv)
time = datetime.datetime.now().strftime("%Y-%m-%d:%H:%M:%S")
software = '%s %s' % (pv.__name__, pv.__version__)
Adjacency.report_events(events_filtered, '%s/sv.bedpe' % args.outdir, sort_by_coord=args.sort_by_coord, header=(software, '%s %s' % (time, cmd)))