def run(args):
bedformat="bed"+str(args.bed_column_number)
dbi=DBI.init(args.genome,"genome")
out=IO.fopen(args.output,"w")
for i in TableIO.parse(IO.fopen(args.input,"r"),bedformat):
print (">",i.id+"_"+args.method,file=out)
print (seq_wrapper(dbi.query(i,method=args.method)),file=out)
def run(args):
schema_template=schema_templates[args.input_format]
SQL_template=insert_templates[args.input_format]
db_filename=args.db
out=IO.fopen(args.output,"w")
if db_filename=="guess":
db_filename=args.input.strip(".gz")+".db"
db_is_new = not os.path.exists(db_filename)
print("Database file : %s"%db_filename,file=out)
with sqlite3.connect(db_filename) as conn:
cursor=conn.cursor()
if db_is_new:
print ('Creating table %s if not exists\n________________________________'%args.table_name,file=out)
S=schema_template.substitute({"table_name":args.table_name})
print (S,file=out)
print ("_______________________________",file=out)
cursor.execute(S)
else:
S=schema_template.substitute({"table_name":args.table_name})
cursor.execute(S)
fin=IO.fopen(args.input,"r")
S1=SQL_template.substitute({"table_name":args.table_name})
print(S1,file=out)
s=TableIO.parse(args.input,"simple")
cursor.executemany(S1,s)
conn.commit()
print("loaded",file=out)
def Main():
'''
IO TEMPLATE
'''
global args,out
args=ParseArg()
fin=IO.fopen(args.input,"r")
out=IO.fopen(args.output,"w")
G2RFP=defaultdict(init);
'''
END OF IO TEMPLATE
'''
m=[0.0 for i in xrange(200)] # 100 to 200
for b in TableIO.parse(IO.fopen(args.db,"r"),"bed6"):
G2RFP[b.chr].append(b)
total_reads=0;
for i0,i in enumerate(TableIO.parse(fin,"bed6")):
spectral=[0 for j in xrange(200)] # 100 to 200
for j in G2RFP[i.chr]:
dis=i.start-j.start
if(dis >=-100 and dis<100): # 50 to 100
spectral[dis+100]+=j.score # 50 to 100
spectral=norm(spectral)
total_reads+=i.score
m=[a*i.score+b for a,b in itertools.izip(spectral,m)]
if i0%100==0:
print("{} processed\r".format(i0),file=sys.stderr)
print("pos\tvalue",file=out);
for i,x in enumerate(m):
print("{}\t{}".format(i,float(x)/total_reads),file=out)
def run(args):
db_filename=args.db
out=IO.fopen(args.output,"w")
if os.path.exists(args.input):
fin=IO.fopen(args.input,"r")
else:
fin=(args.input,)
if not os.path.exists(db_filename):
print("can't find database %s"%db_filename,file=sys.stderr)
exit(1)
print("# Database file : %s"%db_filename,file=out)
with sqlite3.connect(db_filename) as conn:
conn.row_factory=factories[args.db_format]
cursor=conn.cursor()
for i in fin:
i=i.strip()
i=i.strip(" ")
print("# query %s"%i,file=out)
s=template.substitute({"table_name":args.table_name,"name":i})
print("# "+s,file=out)
cursor.execute(s)
r=None
try:
r=cursor.fetchone()
print(r,file=out)
except:
raise
def run(args):
fin=IO.fopen(args.input,"r")
out=IO.fopen(args.output,"w")
r=[]
m=0
ideograms=[]
qr=""
for i in TableIO.parse(fin,sep=","):
if len(i)==1:
a=i[0].split("\t")
if len(a)==2:
if args.query=="all" or args.query==qr:
if m > 0:
ideograms.append({"id":qr,"length":m})
qr=a[1].strip()
else:
if args.query=="all" or args.query==qr:
r.append({"chr":qr,"start":i[0],"length":i[2],"value":i[1]})
m=int(i[0])+int(i[2])
if args.query=="all" or args.query==qr:
if m > 0:
ideograms.append({"id":qr,"length":m})
j={
"ideograms":ideograms,
"tracks":
[
{
"name":args.input,
"type":"bedgraph",
"values":r
}
]
}
print(json.dumps(j,indent=4),file=out)
def run(args):
dbi=DBI.init(args.bw,"bigwig")
out=IO.fopen(args.output,"w")
for i in TableIO.parse(IO.fopen(args.input,"r"),args.format):
ht=[ j for j in dbi.query(i,method=args.method) ]
print("QR",i,file=out)
print("HT",ht,file=out)
def run(args):
fin=IO.fopen(args.input,"r")
out=IO.fopen(args.output,"w")
if args.format=="guess":
args.format=IO.guess_format(args.input)
s=TableIO.parse(args.input,args.format)
for i in s:
print(i,file=out)
def run(local_args):
'''
IO TEMPLATE
'''
global args,out
args=local_args
out=IO.fopen(args.output,"w")
fin=IO.fopen(args.input,"r")
print("# This data was generated by program ",sys.argv[0]," (version: %s)"%VERSION,file=out)
print("# in bam2x ( https://github.com/nimezhu/bam2x )",file=out)
print("# Date: ",time.asctime(),file=out)
print("# The command line is :",file=out)
print("#\t"," ".join(sys.argv),file=out)
gene=DBI.init(args.genetab,"binindex",cls="bed12");
upstream_list=[]
downstream_list=[]
exons_list=[]
introns_list=[]
utr3_list=[]
utr5_list=[]
for g in gene:
upstream_list.append(g.upstream(args.upstream));
downstream_list.append(g.downstream(args.downstream));
for e in g.Exons():
exons_list.append(e)
for i in g.Introns():
introns_list.append(i)
if not (g.utr3() is None):
utr3_list.append(g.utr3())
if not (g.utr5() is None):
utr5_list.append(g.utr5())
upstream=DBI.init(upstream_list,"binindex",cls="bed6")
downstream=DBI.init(downstream_list,"binindex",cls="bed6")
exons=DBI.init(exons_list,"binindex",cls="bed6")
introns=DBI.init(introns_list,"binindex",cls="bed6")
utr3=DBI.init(utr3_list,"binindex",cls="bed6")
utr5=DBI.init(utr5_list,"binindex",cls="bed6")
if args.format=="guess":
args.format=IO.guess_format(args.input)
for (i0,i) in enumerate(TableIO.parse(fin,args.format)):
if i0==0:
if isinstance(i,Bed12):
print("#chr\tstart\tend\tname\tscore\tstrand\tthick_start\tthick_end\titem_rgb\tblock_count\tblock_sizes\tblock_starts\tgene\tupstream\tdownstream\texon\tintron\tutr3\tutr5",file=out)
else:
print("#chr\tstart\tend\tname\tscore\tstrand\tgene\tupstream\tdownstream\texon\tintron\tutr3\tutr5",file=out)
print(i,file=out,end="")
print("\t",toIDs(gene.query(i)),file=out,end="")
print("\t",toIDs(upstream.query(i)),file=out,end="")
print("\t",toIDs(downstream.query(i)),file=out,end="")
print("\t",toIDs(exons.query(i)),file=out,end="")
print("\t",toIDs(introns.query(i)),file=out,end="")
print("\t",toIDs(utr3.query(i)),file=out,end="")
print("\t",toIDs(utr5.query(i)),file=out)
def run(args):
bedformat="bed"+str(args.bed_column_number)
dbi=DBI.init(args.bam,"bam")
out=IO.fopen(args.output,"w")
for i in TableIO.parse(IO.fopen(args.input,"r"),bedformat):
print("QR",i,file=out)
for j in dbi.query(i,method=args.method):
print("HT",j,file=out)
print("",file=out)
def run(args):
bedformat = "bed" + str(args.bed_column_number)
dbi = DBI.init(args.genome, "genome")
out = IO.fopen(args.output, "w")
for i in TableIO.parse(IO.fopen(args.input, "r"), bedformat):
seq = dbi.query(i, method=args.method)
if len(seq) == 0:
continue
print(">{}".format(i.id + "_" + args.method), file=out)
print(seq_wrapper(seq), file=out, end="")
def smart_init(handle,**dict):
'''
test version
'''
if isinstance(handle,str):
fn,ext=os.path.splitext(handle)
if ext==".bam":
dbi=FormatToDBI["bam"]
return dbi(handle,**dict)
elif ext==".gz":
if os.path.isfile(handle+".tbi"):
fn1,ext1=os.path.splitext(fn)
if ext1==".bed":
col_num=IO.get_col_number(handle)
t=ext1+str(col_num)
else:
t=ext1
t=t[1:]
if TableIO.hclass.has_key(t):
return TabixI(handle,cls=t,**dict)
else:
return TabixI(handle,**dict)
else:
fn1,ext1=os.path.splitext(fn)
if ext1==".bed":
col_num=IO.guess(handle)
t=ext1+col_num
else:
t=ext1
t=t[1:]
if TableIO.hclass.has_key(t):
return BinIndexI(handle,cls=t,**dict)
else:
return BinIndexI(handle,**dict)
elif ext==".bed":
col_num=IO.get_col_number(handle)
t=ext+str(col_num)
if TableIO.hclass.has_key(t):
return BinIndexI(handle,cls=t,**dict)
else:
if col_num >=12 :
return BinIndexI(handle,cls="bed12",**dict)
elif col_num>=6:
return BinIndexI(handle,cls="bed6",**dict)
else:
return BinIndexI(handle,cls="bed3",**dict)
else:
if TableIO.hclass.has_key(t):
return BinIndexI(handle,cls=t,**dict)
elif isinstance(handle,file):
return smart_init(handle.name,**dict)
else:
return BinIndex(handle,**dict)
def run(args):
logging.basicConfig(level=logging.INFO)
fin = IO.fopen(args.input, "r")
out = IO.fopen(args.output, "w")
for qr, hits, overlap in iterate(fin):
l = qr.cdna_length()
pileup = [0.0 for i in xrange(l)]
for i in hits:
for j in xrange(i.start, i.stop):
pileup[j] += nh(i)
print("QR\t{id}".format(id=qr.id), file=out)
print("PILEUP\n{value}".format(value=rep(pileup)), file=out)
def run(args):
# logging.basicConfig(level=logging.DEBUG)
dbi = DBI.init(args.bam, "bam")
out = IO.fopen(args.output, "w")
for i in TableIO.parse(IO.fopen(args.input, "r"), "bed12"):
print("QR\t", i, file=out)
for j in dbi.query(i, method="bam1", strand=args.strand):
if compatible_with_transcript(j, i):
print("HT\t{}".format(_translate_to_meta(i, j)), file=out)
elif not args.hit:
print("OP\t{}".format(j), file=out)
print("", file=out)
def run(args):
out=IO.fopen(args.output,"w")
cls=None
if hclass.has_key(args.type):
cls=hclass[args.type]
dbi=DBI.init(args.db,"tabix",cls=cls)
else:
dbi=DBI.init(args.db,"tabix")
for i in TableIO.parse(IO.fopen(args.input,"r"),args.format):
print("QR",i,file=out)
for j,ht in enumerate(dbi.query(i)):
print("HT_{k}\t{ht}".format(k=j+1,ht=ht),file=out)
def run(args):
if os.path.isfile(args.bed + ".tbi"):
dbi = DBI.init(args.bed, "tabix", cls=BED12)
else:
dbi = DBI.init(args.bed, "binindex", cls=BED12)
out = IO.fopen(args.output, "w")
for i in TableIO.parse(IO.fopen(args.input, "r"), "bed12"):
print("QR\t", i, file=out)
for j in dbi.query(i):
if compatible_with_transcript(j, i):
print("HT\t{}".format(_translate_to_meta(i, j)), file=out)
elif not args.hit:
print("OP\t{}".format(j), file=out)
print("", file=out)
def run(args):
fin=IO.fopen(args.input,"r")
out=IO.fopen(args.output,"w")
beds=[i for i in TableIO.parse(fin,"bed12")]
beds.sort()
for i,x in enumerate(iter_cluster(beds)):
id=find_prefix_consensus([i0.id for i0 in x[1]])
strand=find_consensus_strand([i0.strand for i0 in x[1]])
print("REGION\tCL_{index}\t{chr}\t{start}\t{end}\t{id}\t{score}\t{strand}".format(strand=strand,score=len(x[1]),chr=x[1][0].chr,start=x[1][0].start,end=x[0],index=str(i+1),id=id),file=out)
for j,y in enumerate(greedy_iter_compatible_group(x[1])):
print("\tGROUP{j}\t{bed}".format(j=j+1,bed=merge_beds(y,id="CL.{i}_GP.{j}".format(i=i+1,j=j+1))),file=out)
for k,z in enumerate(sorted(y,key= lambda x0:x0.cdna_length(), reverse=True)):
print("\t\tCL.{i}_GP.{j}_TR.{k}\t{l}\t{z}".format(i=i+1,j=j+1,k=k+1,l=z.cdna_length(),z=z),file=out)
def run(args):
fin=IO.fopen(args.input,"r")
outfile=args.input
if not args.sorted:
l = [ i for i in TableIO.parse(fin,args.format) ]
l.sort()
name=splitext(args.input)
outfile = "{name[0]}.sorted{name[1]}".format(name=name)
out = IO.fopen(outfile,"w")
for i in l:
print(i,file=out)
out.close()
format=args.format.translate(None,digits)
tabix_index(outfile,preset=format)
def run(args):
out=IO.fopen(args.output,"w")
if args.annotation=="exon":
for i in TableIO.parse(IO.fopen(args.input,"r"),"bed12"):
for j in i.Exons():
print(j,file=out)
elif args.annotation=="intron":
for i in TableIO.parse(IO.fopen(args.input,"r"),"bed12"):
for j in i.Introns():
print(j,file=out)
elif args.annotation=="cds":
for i in TableIO.parse(IO.fopen(args.input,"r"),"bed12"):
j=i.cds()
if j is not None and j.cdna_length() > 0:
print(j,file=out)
elif args.annotation=="cds":
for i in TableIO.parse(IO.fopen(args.input,"r"),"bed12"):
j=i.cds()
if j is not None and j.cdna_length() > 0:
print(j,file=out)
elif args.annotation=="utr5":
for i in TableIO.parse(IO.fopen(args.input,"r"),"bed12"):
j=i.utr5()
if j is not None and j.cdna_length() > 0:
print(j,file=out)
elif args.annotation=="utr3":
for i in TableIO.parse(IO.fopen(args.input,"r"),"bed12"):
j=i.utr3()
if j is not None and j.cdna_length() > 0:
print(j,file=out)
elif args.annotation=="utr":
for i in TableIO.parse(IO.fopen(args.input,"r"),"bed12"):
j=i.utr5()
if j is not None and j.cdna_length() > 0:
print(j,file=out)
j=i.utr3()
if j is not None and j.cdna_length() > 0:
print(j,file=out)
elif args.annotation=="upstream":
for i in TableIO.parse(IO.fopen(args.input,"r"),"bed12"):
j=i.upstream(args.bp)
print(j,file=out)
elif args.annotation=="downstream":
for i in TableIO.parse(IO.fopen(args.input,"r"),"bed12"):
j=i.downstream(args.bp)
print(j,file=out)
def run(args):
logging.basicConfig(level=logging.DEBUG)
fin=IO.fopen(args.input,"r")
out=IO.fopen(args.output,"w")
bam=DBI.init(args.bam,"bam");
beds=[i for i in TableIO.parse(fin,"bed12")]
beds.sort()
bp=args.bp
print("mapped:{}".format(bam.mapped))
print("unmapped:{}".format(bam.unmapped))
data={}
for i,x in enumerate(iter_cluster(beds)):
print("{}\t{}:{}-{}".format(i+1,x["chr"],x["start"]+1,x["stop"]))
'''
cds=[z.cds() for z in x["beds"] if z.cds()]
utr3=[z.utr3() for z in x["beds"] if z.utr3()]
utr5=[z.utr5() for z in x["beds"] if z.utr5()]
'''
coords = [ up_down_coordinate(gene,args.bp,args.bp) for gene in x["beds"] ]
for j,y in enumerate(coords):
data[y.id]={}
data[y.id]["coord"]=y
data[y.id]["values"]=[0.0 for l in range(y.cdna_length())];
coord_beds = [ _translate(coord,bed) for coord,bed in itertools.izip(coords,x["beds"])]
for j,read in enumerate(bam.query(method="bam1",chr=x["chr"],start=x["start"]-args.bp,stop=x["stop"]+args.bp,strand=args.strand)):
NM=getNM(read) # number of hits
NC=0 # number of compatible
c_coords=[]
for k,coord in enumerate(coords):
if overlap(read,coord) and compatible(read,coord): # don't consider the reads extend out of coords.
NC+=1
c_coords.append(k)
for k,c in enumerate(c_coords):
coord=coords[c]
if read.start < coord.start or read.stop > coord.stop:
start=max(read.start,coord.start)
stop=min(read.stop,coord.stop)
read=read._slice(start,stop)
read_in_coord = _translate(coord,read)
for l in xrange(read_in_coord.start,read_in_coord.stop):
data[coord.id]["values"][l]+=1.0/NC/NM
for j,y in enumerate(coords):
print(data[y.id]["coord"])
print(data[y.id]["values"])
'''
def run(args):
fin=IO.fopen(args.input,"r")
out=IO.fopen(args.output,"w")
if args.format=="guess":
args.format=IO.guess_format(args.input)
s=TableIO.parse(args.input,args.format)
l=[]
for i,x in enumerate(s):
if i/10000==0:
logging.info("reading %s entrys in %s",i,args.input)
l.append(x)
logging.info("begin sorting")
l.sort()
logging.info("sorting done")
for i in l:
print(i,file=out)
logging.info("completed")
def _generate_db(filename,db_filename,table_name):
with sqlite3.connect(db_filename) as conn:
cursor=conn.cursor()
S=schema_t.substitute({"table_name":table_name})
cursor.execute(S)
LOAD_S=insert_t.substitute({"table_name":table_name})
s=TableIO.parse(IO.fopen(filename,"r"),"simple")
cursor.executemany(LOAD_S,s)
conn.commit()
def run(args):
logging.basicConfig(level=logging.INFO)
global bam,out
bam=DBI.init(args.bam,"bam")
fin=IO.fopen(args.input,"r")
out=IO.fopen(args.output,"w")
p=mp.Pool(processes=args.num_cpus)
beds_list=[[] for i in xrange(args.num_cpus)]
for i0,bed in enumerate(TableIO.parse(fin,"bed12")):
beds_list[i0%args.num_cpus].append(bed)
gene_num=i0+1
print("bin_id\tmean\tentropy\treverse_strand_mean\treverse_strand_entropy",file=out)
up_results=p.map(count_flank_star,itertools.izip(beds_list,itertools.repeat(args.bp),itertools.repeat(args.strand),itertools.repeat(True)))
output(up_results,args.bp,gene_num,"UP")
results = p.map(count_list_star,itertools.izip(beds_list,itertools.repeat(args.bin_num),itertools.repeat(args.strand)))
output(results,args.bin_num,gene_num,"TR")
down_results=p.map(count_flank_star,itertools.izip(beds_list,itertools.repeat(args.bp),itertools.repeat(args.strand),itertools.repeat(False)))
output(down_results,args.bp,gene_num,"DN")
def run(local_args):
logging.basicConfig(level=logging.WARNING)
global args,out,dbi_bam,g, MIN_INTRON_LENGTH, MIN_SPLICING_SITES_SCORE, MIN_FPK_RATIO,query_num
MIN_INTRON_LENGTH=10
MIN_SPLICING_SITES_SCORE=2
'''
IO TEMPLATE
'''
'''
mySorts={ 0:sort_by_intron_and_abundance,
1:sort_by_intron,
2:sort_by_abundance
}
'''
args=local_args
#print "debug:",args.report_seq
MIN_FPK_RATIO=args.min_uniq_fpk_increase #TO TEST
fin=IO.fopen(args.input,"r")
out=IO.fopen(args.output,"w")
'''
END OF IO TEMPLATE
'''
print >>out,"# This data was generated by program ",sys.argv[0]," (version: %s)"%VERSION,
print >>out,"in bam2x ( https://github.com/nimezhu/bam2x )"
print >>out,"# Date: ",time.asctime()
print >>out,"# The command line is :"
print >>out,"#\t"," ".join(sys.argv)
# header=["chr","start","end","id","score","strand","seq"];
# dbi_splicing_sites=DBI.init(args.splicing_sites,"tabix",tabix="metabed",header=header);
#if args.format=="guess":
# args.format=IO.guess_format(args.input)
reader=TableIO.parse(fin,args.format)
query_list=[]
query_lists=[[] for i in range(args.num_cpus)]
query_num=0
for i,x in enumerate(reader):
query_lists[i%args.num_cpus].append(x)
query_num=i+1
#querys(query_lists[0]) #DEBUG
pool=Pool(processes=args.num_cpus)
results=pool.map(querys,query_lists)
#print results
output(results)
def run(args):
logging.basicConfig(level=logging.DEBUG)
db_filename=args.translator
t_name,t_ext=splitext(args.translator)
'''
test if it is db file
generate db file if it doesn't exists.
'''
if t_ext!="db":
#possible_db=args.translator.strip("\\.gz")+".db"
possible_db=args.translator+".db"
print(possible_db)
if os.path.exists(possible_db):
db_filename=possible_db
else:
_generate_db(args.translator,possible_db,args.table_name)
db_filename=possible_db
'''
query db file
'''
out=IO.fopen(args.output,"w")
with sqlite3.connect(db_filename) as conn:
conn.row_factory=lambda conn,x: Bed12._make(Bed12._types(x[1:]))
cursor=conn.cursor()
for i in TableIO.parse(IO.fopen(args.input,"r"),"bed"):
s=template.substitute({"table_name":args.table_name,"name":i.chr.strip()})
print(s)
cursor.execute(s)
gene=None
try:
gene=cursor.fetchone()
logging.debug(i)
logging.debug(i.cdna_length())
logging.debug(gene)
logging.debug(gene.cdna_length())
except:
raise
logging.warning("can't find gene %s"%i.chr)
continue
assert gene.cdna_length() > i.cdna_length()
print(reverse_translate(gene,i),file=out)
def run(args_local):
'''
IO TEMPLATE
'''
global args,out,exon_cutoff,intron_cutoff,hasGenome
args=args_local
dbi=DBI.init(args.input,"bam")
out=IO.fopen(args.output,"w")
hasGenome=False
if args.genome:
hasGenome=True
'''
END OF IO TEMPLATE
'''
print >>out,"# This positive_data was generated by program ",sys.argv[1]," (version: %s)"%VERSION,
print >>out,"in bam2x ( https://github.com/nimezhu/bam2x )"
print >>out,"# Date: ",time.asctime()
print >>out,"# The command line is :"
print >>out,"#\t"," ".join(sys.argv)
chrs=[]
lengths=[]
for i in dbi.bamfiles[0].references:
chrs.append(i)
for i in dbi.bamfiles[0].lengths:
lengths.append(i)
p=mp.Pool(processes=args.num_cpus)
coverage_bedgraphs=p.map(process_chrom,chrs)
bedgraphs=[]
coverages=[]
for i in range(len(chrs)):
bedgraphs.append(coverage_bedgraphs[i][1])
coverages.append(coverage_bedgraphs[i][0])
s=0.0 # 1000.0
l=long(0)
for i in range(len(chrs)):
s+=coverages[i]
l+=lengths[i]
l=l*2 # Double Strand
coverage=s/l*1000.0
threshold=1
while 1:
if prob.poisson_cdf(threshold,coverage,False) < args.pvalue: break
threshold+=1
exon_cutoff=threshold
intron_cutoff=2 #TODO revise it
print >>out,"# MEAN COVERAGE:",coverage
print >>out,"# EXON COVERAGE CUTOFF:",exon_cutoff
#call_peaks(bedgraphs[0],1) #debug
peaks=p.map(call_peaks_star,itertools.izip(chrs,bedgraphs,itertools.repeat(exon_cutoff)))
output(chrs,peaks)
def run(args):
#logging.basicConfig(level=logging.DEBUG)
dbi=DBI.init(args.bam,"bam")
mapped=dbi.mapped
out=IO.fopen(args.output,"w")
print("Gene\tRPKM",file=out);
for i in TableIO.parse(IO.fopen(args.input,"r"),"bed12"):
print(i.id,"\t",end="",file=out)
s=0.0
l=i.cdna_length()
if args.uniq:
for j in dbi.query(i,method="bam1",strand=args.strand,uniq=args.uniq):
if compatible_with_transcript(j,i):
s+=1.0
else:
for j in dbi.query(i,method="bam1",strand=args.strand,uniq=args.uniq):
if compatible_with_transcript(j,i):
(nh,_,_)=j.itemRgb.split(",")
nh=int(nh)
s+=1.0/nh
rpkm=float(s)*(1000000.0/mapped)*(1000.0/float(l))
print(rpkm,file=out)
def parse_simple(handle,**dict):
sep="\t"
if dict.has_key("sep"):
sep=dict["sep"]
if isinstance(handle,str):
try:
handle=IO.fopen(handle,"r")
for i in csv.reader(handle,delimiter=sep):
if i[0].strip()[0]=="#": continue
yield tuple(i)
handle.close()
except IOError as e:
print >>sys.stderr,"I/O error({0}): {1}".format(e.errno, e.strerror)
else:
try:
for i in csv.reader(handle,delimiter=sep):
if i[0].strip()[0]=="#": continue
yield tuple(i)
except:
raise
def __init__(self,file,**dict):
'''
Wrapped in bam2x.DBI.init()
BinIndex(file,cls=inherited_namedtuplecls)
inherited_namedtuplecls should have _make and _types functino
or
assuming the entry in container is already formatted
BinIndex(container)
'''
if isinstance(file,str):
file=csv.reader(IO.fopen(file,"r"),delimiter="\t")
if dict.has_key("cls"):
cls=dict["cls"]
if isinstance(cls,str):
if hclass.has_key(cls):
cls=hclass[cls]
else:
print >>sys.stderr,"UNKNOWN FORMAT %s IN BININDEX DATA STRUCT"%cls
self.data=binindex(file,cls=cls)
else:
self.data=binindex(file)
def run(args):
logging.basicConfig(level=logging.INFO)
up=args.up
down=args.down
bp_num=up+down
offset=-up
bam=DBI.init(args.bam,"bam")
fin=IO.fopen(args.input,"r")
out=IO.fopen(args.output,"w")
bin_sum=[0 for i in xrange(bp_num)]
bin_e=[0.0 for i in xrange(bp_num)]
bin_dis=[[] for i in xrange(bp_num)]
for i0,bed in enumerate(TableIO.parse(fin,args.format)):
bed_bin=[0 for i in xrange(bp_num)]
if args.tts:
pos=bed.tts()
else:
pos=bed.tss()
pos_flank=get_flank_region(pos,up,down)
for read in bam.query(pos_flank,"bam1",strand="read1"):
a=translate_coordinates(pos,read)
#print(a,file=out)
for e in a.Exons():
#print(e,file=out)
start=e.start-offset
end=e.stop-offset
if start < 0: start=0
if end > bp_num: end=bp_num
for j in xrange(start,end):
bed_bin[j]+=1
for i in xrange(bp_num):
bin_sum[i]+=bed_bin[i]
bin_dis[i].append(bed_bin[i])
bed_num=i0+1
for i in xrange(bp_num):
bin_e[i]=gini_coefficient(bin_dis[i])
if args.tts:
print("pos_to_tts\taggregation_mean\tgini_coefficient",file=out)
else:
print("pos_to_tss\taggregation_mean\tgini_coefficient",file=out)
for i in xrange(bp_num):
print("{bin}\t{aggregation}\t{E}".format(bin=i+offset,aggregation=float(bin_sum[i])/bed_num,E=bin_e[i]),file=out)
try:
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt
matplotlib.rcParams.update({'font.size':9})
ax1=plt.subplot2grid((7,1),(6,0))
plt.ylabel('gini coeffecient')
plt.fill_between(range(-up,down),bin_e,color="r",alpha=0.2,y2=0)
ax1.set_ylim(0,1)
ax1.set_xlim(-up,down)
ax1.axes.get_xaxis().set_visible(False)
plt.axvline(x=0,linewidth=1, color='y')
ax2=plt.subplot2grid((7,1),(0,0),rowspan=5)
ax2.set_xlim(-up,down)
plt.plot(range(-up,down),[float(i)/bed_num for i in bin_sum])
plt.ylabel('mean coverage')
if args.tts:
plt.xlabel('pos to tts (bp)')
else:
plt.xlabel('pos to tss (bp)')
plt.axvline(x=0,linewidth=1, color='y')
plt.grid(True)
plt.savefig(args.output+".png")
except:
pass
def run(args):
logging.basicConfig(level=logging.INFO)
def process():
if len(buff)==1: return 0
max_score=0.0
total_score=0.0
e=[]
for i in buff:
total_score+=i.score
e.append(i.score)
e=[i/total_score for i in e]
gini=gini_coefficient(e)
if total_score < args.min_reads_number:
return 0
record={}
meta=BED6(buff[0].chr,buff[0].start,buff[-1].stop,args.prefix+"."+str(group_id),total_score,buff[0].strand)
peak=max(buff,key=lambda x:x.score)
record["peak"]=peak._replace(score=peak.score/total_score)
record["meta"]=meta._replace(strand=peak.strand)
record["gini"]=gini
records.append(record)
return 1
def simple_output():
print("# formats: bayes_prob_model2, gini, [ region bed, score is total reads], [peak bed , score is proportion ]",file=out)
for i,x in enumerate(records):
print("{p2}\t{gini}\t{meta}\t".format(p2=p2[i],meta=x["meta"],gini=x["gini"]),end="",file=out)
print(x["peak"],file=out)
def bed12_output():
print("# formats: bed12 , [R,G,B] are corresponding to [ TTS_GINI_PVALUE*200, TSS_GINI_PALUE*200, PROPORTION_OF_PEAK*200 ]",file=out)
for i,x in enumerate(records):
if args.tts:
g=0
r=int(p2[i]*200)
else:
g=int(p2[i]*200)
r=0
b=int(x["gini"]*200)
if p2[i]>0.5:
meta=x["meta"]._replace(id=x["meta"].id+".end")
else:
meta=x["meta"]
rgb="{r},{g},{b}".format(r=r,g=g,b=b)
print("{bed6}\t{thickStart}\t{thickEnd}\t{itemRgb}\t{blockCount}\t{blockSizes}\t{blockStarts}".format(bed6=meta,thickStart=x["peak"].start,thickEnd=x["peak"].end,itemRgb=rgb,blockSizes=x["meta"].stop-x["meta"].start,blockCount=1,blockStarts=0),file=out)
records=[]
GAP=args.gap
fin=IO.fopen(args.input,"r")
out=IO.fopen(args.output,"w")
iterator=TableIO.parse(fin,"bed6")
last=iterator.next()
last_stop=last.stop
group_id=0
buff=[last]
last_chr=last.chr
for x,i in enumerate(iterator):
if x%10000==0: logging.info("processing {x} reads".format(x=x));
if i.chr!=last_chr or i.start-last_stop > GAP:
group_id+=process()
buff=[i]
last_chr=i.chr
last_stop=i.stop
else:
buff.append(i)
if i.stop>last_stop:
last_stop=i.stop
process()
gini=array([i["gini"] for i in records])
model=fit_two_peaks_EM(gini)
p2=bayes_p2(gini,model)
print("# Date: ",time.asctime(),file=out)
print("# Program Version ",VERSION,file=out)
print("# The command line is :",file=out)
print("#\t"," ".join(sys.argv),file=out)
print("# learning model:",file=out)
print("#",model_str(model),file=out)
#simple_output()
bed12_output()
def run(args):
fin=IO.fopen(args.input,"r")
out=IO.fopen(args.output,"w")
for i in TableIO.parse(fin,"bed12"):
print(remove_small_introns(i,args.cutoff),file=out)