def generate_transcript_counts(data):
"""Generate counts per transcript and per exon from an alignment"""
data["count_file"] = featureCounts.count(data)
if dd.get_transcriptome_align(data):
# to create a disambiguated transcriptome file realign with bowtie2
if dd.get_disambiguate(data):
logger.info("Aligning to the transcriptome with bowtie2 using the "
"disambiguated reads.")
bam_path = data["work_bam"]
fastq_paths = alignprep._bgzip_from_bam(
bam_path,
data["dirs"],
data,
is_retry=False,
output_infix='-transcriptome')
if len(fastq_paths) == 2:
file1, file2 = fastq_paths
else:
file1, file2 = fastq_paths[0], None
ref_file = dd.get_ref_file(data)
data = bowtie2.align_transcriptome(file1, file2, ref_file, data)
else:
file1, file2 = dd.get_input_sequence_files(data)
if not dd.get_transcriptome_bam(data):
ref_file = dd.get_ref_file(data)
logger.info(
"Transcriptome alignment was flagged to run, but the "
"transcriptome BAM file was not found. Aligning to the "
"transcriptome with bowtie2.")
data = bowtie2.align_transcriptome(file1, file2, ref_file, data)
data = spikein.counts_spikein(data)
return [[data]]
def generate_transcript_counts(data):
"""Generate counts per transcript and per exon from an alignment"""
data["count_file"] = featureCounts.count(data)
if dd.get_fusion_mode(data, False):
oncofuse_file = oncofuse.run(data)
if oncofuse_file:
data = dd.set_oncofuse_file(data, oncofuse_file)
if dd.get_transcriptome_align(data) and not dd.get_transcriptome_bam(data):
file1, file2 = None, None
if dd.get_disambiguate(data):
bam_path = data["work_bam"]
fastq_paths = alignprep._bgzip_from_bam(bam_path, data["dirs"], data["config"], is_retry=False, output_infix='-transcriptome')
if len(fastq_paths) == 2:
file1, file2 = fastq_paths
else:
file1, file2 = fastq_paths[0], None
else:
file1, file2 = dd.get_input_sequence_files(data)
ref_file = dd.get_ref_file(data)
logger.info("Transcriptome alignment was flagged to run, but the "
"transcriptome BAM file was not found. Aligning to the "
"transcriptome with bowtie2.")
data = bowtie2.align_transcriptome(file1, file2, ref_file, data)
return [[data]]
def _update_data(align_file, out_dir, names, data):
data = dd.set_work_bam(data, align_file)
data = dd.set_align_bam(data, align_file)
if dd.get_transcriptome_align(data) and not is_transcriptome_broken(data):
transcriptome_file = _move_transcriptome_file(out_dir, names)
data = dd.set_transcriptome_bam(data, transcriptome_file)
return data
def generate_transcript_counts(data):
"""Generate counts per transcript and per exon from an alignment"""
data["count_file"] = featureCounts.count(data)
if dd.get_fusion_mode(data, False):
oncofuse_file = oncofuse.run(data)
if oncofuse_file:
data = dd.set_oncofuse_file(data, oncofuse_file)
if dd.get_transcriptome_align(data) and not dd.get_transcriptome_bam(data):
file1, file2 = None, None
if dd.get_disambiguate(data):
bam_path = data["work_bam"]
fastq_paths = alignprep._bgzip_from_bam(
bam_path,
data["dirs"],
data["config"],
is_retry=False,
output_infix='-transcriptome')
if len(fastq_paths) == 2:
file1, file2 = fastq_paths
else:
file1, file2 = fastq_paths[0], None
else:
file1, file2 = dd.get_input_sequence_files(data)
ref_file = dd.get_ref_file(data)
logger.info("Transcriptome alignment was flagged to run, but the "
"transcriptome BAM file was not found. Aligning to the "
"transcriptome with bowtie2.")
data = bowtie2.align_transcriptome(file1, file2, ref_file, data)
return [[data]]
def _update_data(align_file, out_dir, names, data):
data = dd.set_work_bam(data, align_file)
data = dd.set_align_bam(data, align_file)
if dd.get_transcriptome_align(data) and not is_transcriptome_broken():
transcriptome_file = _move_transcriptome_file(out_dir, names)
data = dd.set_transcriptome_bam(data, transcriptome_file)
return data
def align(fastq_file, pair_file, ref_file, names, align_dir, data):
max_hits = 10
srna = True if data["analysis"].lower().startswith("smallrna-seq") else False
srna_opts = ""
if srna:
max_hits = 1000
srna_opts = "--alignIntronMax 1"
config = data["config"]
out_prefix = os.path.join(align_dir, dd.get_lane(data))
out_file = out_prefix + "Aligned.out.sam"
out_dir = os.path.join(align_dir, "%s_star" % dd.get_lane(data))
if not ref_file:
logger.error("STAR index not found. We don't provide the STAR indexes "
"by default because they are very large. You can install "
"the index for your genome with: bcbio_nextgen.py upgrade "
"--aligners star --genomes genome-build-name --data")
sys.exit(1)
final_out = os.path.join(out_dir, "{0}.bam".format(names["sample"]))
if file_exists(final_out):
data = _update_data(final_out, out_dir, names, data)
return data
star_path = config_utils.get_program("STAR", config)
fastq_files = " ".join([fastq_file, pair_file]) if pair_file else fastq_file
num_cores = dd.get_num_cores(data)
gtf_file = dd.get_gtf_file(data)
safe_makedir(align_dir)
cmd = ("{star_path} --genomeDir {ref_file} --readFilesIn {fastq_files} "
"--runThreadN {num_cores} --outFileNamePrefix {out_prefix} "
"--outReadsUnmapped Fastx --outFilterMultimapNmax {max_hits} "
"--outStd SAM {srna_opts} "
"--outSAMunmapped Within --outSAMattributes %s " % " ".join(ALIGN_TAGS))
cmd += _add_sj_index_commands(fastq_file, ref_file, gtf_file)
cmd += " --readFilesCommand zcat " if is_gzipped(fastq_file) else ""
cmd += _read_group_option(names)
fusion_mode = utils.get_in(data, ("config", "algorithm", "fusion_mode"), False)
if fusion_mode:
cmd += " --chimSegmentMin 15 --chimJunctionOverhangMin 15"
strandedness = utils.get_in(data, ("config", "algorithm", "strandedness"),
"unstranded").lower()
if strandedness == "unstranded" and not srna:
cmd += " --outSAMstrandField intronMotif "
if dd.get_transcriptome_align(data) and not is_transcriptome_broken(data):
cmd += " --quantMode TranscriptomeSAM "
with file_transaction(data, final_out) as tx_final_out:
cmd += " | " + postalign.sam_to_sortbam_cl(data, tx_final_out)
run_message = "Running STAR aligner on %s and %s" % (fastq_file, ref_file)
do.run(cmd.format(**locals()), run_message, None)
data = _update_data(final_out, out_dir, names, data)
return data
def align(fastq_file, pair_file, ref_file, names, align_dir, data):
max_hits = 10
srna = True if data["analysis"].lower().startswith("smallrna-seq") else False
srna_opts = ""
if srna:
max_hits = 1000
srna_opts = "--alignIntronMax 1"
config = data["config"]
out_prefix = os.path.join(align_dir, dd.get_lane(data))
out_file = out_prefix + "Aligned.out.sam"
out_dir = os.path.join(align_dir, "%s_star" % dd.get_lane(data))
if not ref_file:
logger.error("STAR index not found. We don't provide the STAR indexes "
"by default because they are very large. You can install "
"the index for your genome with: bcbio_nextgen.py upgrade "
"--aligners star --genomes genome-build-name --data")
sys.exit(1)
final_out = os.path.join(out_dir, "{0}.bam".format(names["sample"]))
if file_exists(final_out):
data = _update_data(final_out, out_dir, names, data)
return data
star_path = config_utils.get_program("STAR", config)
fastq = " ".join([fastq_file, pair_file]) if pair_file else fastq_file
num_cores = config["algorithm"].get("num_cores", 1)
safe_makedir(align_dir)
cmd = ("{star_path} --genomeDir {ref_file} --readFilesIn {fastq} "
"--runThreadN {num_cores} --outFileNamePrefix {out_prefix} "
"--outReadsUnmapped Fastx --outFilterMultimapNmax {max_hits} "
"--outStd SAM {srna_opts} "
"--outSAMunmapped Within --outSAMattributes %s" % " ".join(ALIGN_TAGS))
cmd = cmd + " --readFilesCommand zcat " if is_gzipped(fastq_file) else cmd
cmd += _read_group_option(names)
fusion_mode = utils.get_in(data, ("config", "algorithm", "fusion_mode"), False)
if fusion_mode:
cmd += " --chimSegmentMin 15 --chimJunctionOverhangMin 15"
strandedness = utils.get_in(data, ("config", "algorithm", "strandedness"),
"unstranded").lower()
if strandedness == "unstranded" and not srna:
cmd += " --outSAMstrandField intronMotif "
if dd.get_transcriptome_align(data) and not is_transcriptome_broken():
cmd += " --quantMode TranscriptomeSAM "
with file_transaction(data, final_out) as tx_final_out:
cmd += " | " + postalign.sam_to_sortbam_cl(data, tx_final_out)
run_message = "Running STAR aligner on %s and %s" % (fastq_file, ref_file)
do.run(cmd.format(**locals()), run_message, None)
data = _update_data(final_out, out_dir, names, data)
return data
