def is_tRNA(clus_obj, out_dir, args):
"""
Iterates through cluster precursors to predict sRNA types
"""
ref = os.path.abspath(args.reference)
utils.safe_dirs(out_dir)
for nc in clus_obj[0]:
c = clus_obj[0][nc]
loci = c['loci']
out_fa = "cluster_" + nc
if loci[0][3] - loci[0][2] < 500:
with make_temp_directory() as tmpdir:
os.chdir(tmpdir)
get_loci_fasta({loci[0][0]: [loci[0][0:5]]}, out_fa, ref)
summary_file, str_file = _run_tRNA_scan(out_fa)
if "predictions" not in c:
c['predictions'] = {}
c['predictions']['tRNA'] = _read_tRNA_scan(summary_file)
score = _read_tRNA_scan(summary_file)
logger.debug(score)
shutil.move(summary_file, op.join(out_dir, summary_file))
shutil.move(str_file, op.join(out_dir, str_file))
else:
c['errors'].add("precursor too long")
clus_obj[0][nc] = c
return clus_obj
def make_predictions(clus_obj, out_dir, args):
"""
Iterates through cluster precursors to predict sRNA types
"""
ref = os.path.abspath(args.reference)
utils.safe_dirs(out_dir)
for nc in clus_obj[0]:
c = clus_obj[0][nc]
loci = c['loci']
out_fa = "cluster_" + nc
if loci[0][3] - loci[0][2] < 500:
with make_temp_directory() as tmpdir:
os.chdir(tmpdir)
get_loci_fasta({loci[0][0]: [loci[0][0:5]]}, out_fa, ref)
summary_file, str_file = _run_tRNA_scan(out_fa)
if "predictions" not in c:
c['predictions'] = {}
c['predictions']['tRNA'] = _read_tRNA_scan(summary_file)
score = _read_tRNA_scan(summary_file)
logger.debug(score)
shutil.move(summary_file, op.join(out_dir, summary_file))
shutil.move(str_file, op.join(out_dir, str_file))
else:
c['errors'].add("precursor too long")
clus_obj[0][nc] = c
return clus_obj
def make_profile(data, out_dir, args):
"""
Make data report for each cluster
"""
safe_dirs(out_dir)
main_table = []
header = ['id', 'ann']
n = len(data[0])
bar = ProgressBar(maxval=n).start()
bar.update(0)
for itern, c in enumerate(data[0]):
bar.update(itern)
logger.debug("creating cluser: {}".format(c))
safe_dirs(os.path.join(out_dir, c))
valid, ann, pos_structure = _single_cluster(c, data, os.path.join(out_dir, c, "maps.tsv"), args)
data[0][c].update({'profile': pos_structure})
loci = data[0][c]['loci']
logger.debug("precursor_sequence")
data[0][c]['precursor'] = {"seq": precursor_sequence(loci[0][0:5], args.ref)}
logger.debug("parse alignments")
data[0][c]['precursor']["colors"] = list(_parse(data[0][c]['profile'], data[0][c]['precursor']["seq"]))
logger.debug("update rnafold")
data[0][c]['precursor'].update(run_rnafold(data[0][c]['precursor']['seq']))
return data
def make_profile(data, out_dir, args):
"""
Make data report for each cluster
"""
safe_dirs(out_dir)
main_table = []
header = ['id', 'ann']
n = len(data[0])
bar = ProgressBar(maxval=n)
bar.start()
bar.update(0)
for itern, c in enumerate(data[0]):
bar.update(itern)
logger.debug("creating cluser: {}".format(c))
safe_dirs(os.path.join(out_dir, c))
valid, ann, pos_structure = _single_cluster(
c, data, os.path.join(out_dir, c, "maps.tsv"), args)
data[0][c].update({'profile': pos_structure})
loci = data[0][c]['loci']
data[0][c]['precursor'] = {
"seq": precursor_sequence(loci[0][0:5], args.ref)
}
data[0][c]['precursor']["colors"] = _parse(
data[0][c]['profile'], data[0][c]['precursor']["seq"])
data[0][c]['precursor'].update(
run_rnafold(data[0][c]['precursor']['seq']))
return data
def run_coral(clus_obj, out_dir, args):
"""
Run some CoRaL modules to predict small RNA function
"""
if not args.bed:
raise ValueError("This module needs the bed file output from cluster subcmd.")
workdir = op.abspath(op.join(args.out, 'coral'))
safe_dirs(workdir)
bam_in = op.abspath(args.bam)
bed_in = op.abspath(args.bed)
reference = op.abspath(args.ref)
with chdir(workdir):
bam_clean = coral.prepare_bam(bam_in, bed_in)
out_dir = op.join(workdir, "regions")
safe_dirs(out_dir)
prefix = "seqcluster"
loci_file = coral.detect_regions(bam_clean, bed_in, out_dir, prefix)
coral.create_features(bam_clean, loci_file, reference, out_dir)
def run_coral(clus_obj, out_dir, args):
"""
Run some CoRaL modules to predict small RNA function
"""
if not args.bed:
raise ValueError(
"This module needs the bed file output from cluster subcmd.")
workdir = op.abspath(op.join(args.out, 'coral'))
safe_dirs(workdir)
bam_in = op.abspath(args.bam)
bed_in = op.abspath(args.bed)
reference = op.abspath(args.ref)
with chdir(workdir):
bam_clean = coral.prepare_bam(bam_in, bed_in)
out_dir = op.join(workdir, "regions")
safe_dirs(out_dir)
prefix = "seqcluster"
loci_file = coral.detect_regions(bam_clean, bed_in, out_dir, prefix)
coral.create_features(bam_clean, loci_file, reference, out_dir)
def predictions(args):
"""
Create predictions of clusters
"""
logger.info(args)
logger.info("Reading sequences")
out_file = os.path.abspath(os.path.splitext(args.json)[0] + "_prediction.json")
data = load_data(args.json)
out_dir = os.path.abspath(safe_dirs(os.path.join(args.out, "predictions")))
logger.info("Make predictions")
data = is_tRNA(data, out_dir, args)
if args.coral:
logger.info("Make CoRaL predictions")
run_coral(data, out_dir, args)
write_data(data[0], out_file)
logger.info("Done")
def predictions(args):
"""
Create predictions of clusters
"""
logger.info(args)
logger.info("reading sequeces")
out_file = os.path.abspath(os.path.splitext(args.json)[0] + "_prediction.json")
data = load_data(args.json)
out_dir = os.path.abspath(safe_dirs(os.path.join(args.out, "predictions")))
logger.info("make predictions")
data = is_tRNA(data, out_dir, args)
if args.coral:
logger.info("make CoRaL predictions")
run_coral(data, out_dir, args)
write_data(data[0], out_file)
logger.info("Done")
def make_database(data, name="seqcluster.db", out_dir="database"):
out_dir = safe_dirs(out_dir)
op.abspath(out_dir)
con = _create_db(op.join(out_dir, name))
_insert_data(con, data)
_close(con)
def make_database(data, name="seqcluster.db", out_dir="database"):
out_dir = safe_dirs(out_dir)
op.abspath(out_dir)
con = _create_db(op.join(out_dir, name))
_insert_data(con, data)
_close(con)