def Readanno(filename, annoglb, genome):
glannot = glload(annoglb)
allelement = set(glannot['annot'])
if genome in ['mm10']:
chr_list = ['chr' + str(i)
for i in range(1, 20)] + ['chrX', 'chrY', 'chrM']
elif genome in ['hg38']:
chr_list = ['chr' + str(i)
for i in range(1, 23)] + ['chrX', 'chrY', 'chrM']
return (allelement, chr_list, annoglb, glannot)
def Readanno(filename, annoglb, genome):
glannot = glload(annoglb)
allelement = set(glannot['annot'])
# if genome in ['mm10']:
# chr_list = ['chr'+ str(i) for i in range(1,20) ] + [ 'chrX','chrY', 'chrM' ]
# elif genome in ['hg38']:
# chr_list = ['chr'+ str(i) for i in range(1,23) ] + [ 'chrX','chrY', 'chrM' ]
chr_list = list(set([k['chr'] for k in glannot['loc']
])) #this is useful for costume chromsome
return (allelement, chr_list, annoglb, glannot)
def getanno(filename, genefile, tefile, genome, mode):
form = {
'force_tsv': True,
'loc': 'location(chr=column[0], left=column[1], right=column[2])',
'annot': 3
}
if genefile == 'default' and tefile == 'default':
if genome == 'mm10':
chr_list = ['chr' + str(i)
for i in range(1, 20)] + ['chrX', 'chrY', 'chrM']
if mode == 'exclusive':
if not os.path.exists('mm10.exclusive.glb'):
logging.error(
"Did not find the annotation index mm10.exclusive.glb, you can download it from scTE github (www....) or either give the annotation with -te and -gene option \n"
)
sys.exit(1)
all_annot = 'mm10.exclusive.glb'
allelement = set(glload(all_annot)['annot'])
elif mode == 'inclusive':
if not os.path.exists('mm10.inclusive.glb'):
logging.error(
"Did not find the annotation index mm10.inclusive.glb, you can download it from scTE github (www....) or either give the annotation with -te and -gene option \n"
)
sys.exit(1)
all_annot = 'mm10.inclusive.glb'
allelement = set(glload(all_annot)['annot'])
elif genome == 'hg38':
chr_list = ['chr' + str(i)
for i in range(1, 23)] + ['chrX', 'chrY', 'chrM']
if mode == 'exclusive':
if not os.path.exists('hg38.exclusive.glb'):
logging.error(
"Did not find the annotation index hg38.exclusive.glb, you can download it from scTE github (www....) or either give the annotation with -te and -gene option \n"
)
sys.exit(1)
all_annot = 'hg38.exclusive.glb'
allelement = set(glload(all_annot)['annot'])
elif mode == 'inclusive':
if not os.path.exists('hg38.inclusive.glb'):
logging.error(
"Did not find the annotation index hg38.inclusive.glb, you can download it from scTE github (www....) or either give the annotation with -te and -gene option \n"
)
sys.exit(1)
all_annot = 'hg38.inclusive.glb'
allelement = set(glload(all_annot)['annot'])
else:
if genome in ['hg38']:
chr_list = ['chr' + str(i)
for i in range(1, 23)] + ['chrX', 'chrY', 'chrM']
elif genome in ['mm10']:
chr_list = ['chr' + str(i)
for i in range(1, 20)] + ['chrX', 'chrY', 'chrM']
if not os.path.isfile(tefile):
logging.error("No such file: %s !\n" % (tefile))
sys.exit(1)
if not os.path.isfile(genefile):
logging.error("No such file: %s !\n" % (genefile))
sys.exit(1)
all_annot = annoGtf(filename,
genefile=genefile,
tefile=tefile,
mode=mode)
allelement = set(glload(all_annot)['annot'])
return (allelement, chr_list, all_annot)
def align(chr, filename, all_annot, glannot, whitelist, CB):
'''
**Purpose**
For each read, align it to the index and assign a TE, gene.
This is the speed critical part.
'''
s1 = time.time()
chr = 'chr' + chr
if not os.path.exists('%s_scTEtmp/o3' % filename):
os.system('mkdir -p %s_scTEtmp/o3' % filename)
if not glannot: # Load separately for the multicore pipeline, share the index for the single core pipeline
glannot = glload(all_annot)
# Only keep the glbase parts we need.
buckets = glannot.buckets[chr.replace('chr', '')]
all_annot = glannot.linearData
oh = gzip.open('%s_scTEtmp/o2/%s.%s.bed.gz' % (filename, filename, chr),
'rt')
res = {}
for line in oh:
t = line.strip().split('\t')
barcode = t[3]
if barcode not in whitelist:
continue
if barcode not in res:
res[barcode] = defaultdict(int)
#chrom = t[0].replace('chr', '') # Don't need as each align is already split for each chrom;
left = int(t[1])
rite = int(t[2])
#loc = location(chr=chrom, left=left, right=rite)
left_buck = ((left - 1) // 10000) * 10000
right_buck = ((rite) // 10000) * 10000
buckets_reqd = range(left_buck, right_buck + 10000, 10000)
if buckets_reqd:
loc_ids = set()
loc_ids_update = loc_ids.update
# get the ids reqd.
[
loc_ids_update(buckets[buck]) for buck in buckets_reqd
if buck in buckets
]
result = [
all_annot[index]['annot'] for index in loc_ids
if (rite >= all_annot[index]['loc'].loc['left']
and left <= all_annot[index]['loc'].loc["right"])
]
if result:
for gene in result:
res[barcode][gene] += 1
oh.close()
oh = gzip.open('%s_scTEtmp/o3/%s.%s.bed.gz' % (filename, filename, chr),
'wt')
for bc in sorted(res):
for gene in sorted(res[bc]):
oh.write('%s\t%s\t%s\n' % (bc, gene, res[bc][gene]))
oh.close()
