def querys(listx):
dbi_bam=DBI.init(args.bam,"bam")
genome=DBI.init(args.genome,"genome")
results=[]
for x in listx:
results.append(query(x,dbi_bam,genome))
return results
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):
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):
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):
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 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):
# 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_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 process_chrom(chrom):
local_dbi=DBI.init(args.input,"bam")
retv=list()
intron_retv=list()
a=[]
#positive_data=TuringSortingArray(None,500)
positive_data=TuringSortingArray()
negative_data=TuringSortingArray()
positive_intron_data=TuringSortingArray()
negative_intron_data=TuringSortingArray()
for i in local_dbi.query(chrom,method="bam1",strand=args.strand):
if i.strand=="+" or i.strand==".":
positive_data.append(TuringCode(i.start,cb.ON))
positive_data.append(TuringCode(i.stop,cb.OFF))
for j in i.Exons():
positive_data.append(TuringCode(j.start,cb.BLOCKON))
positive_data.append(TuringCode(j.stop,cb.BLOCKOFF))
for j in i.Introns():
positive_intron_data.append(TuringCode(j.start,cb.BLOCKON))
positive_intron_data.append(TuringCode(j.stop,cb.BLOCKOFF))
else:
negative_data.append(TuringCode(i.start,cb.ON))
negative_data.append(TuringCode(i.stop,cb.OFF))
for j in i.Exons():
negative_data.append(TuringCode(j.start,cb.BLOCKON))
negative_data.append(TuringCode(j.stop,cb.BLOCKOFF))
for j in i.Introns():
negative_intron_data.append(TuringCode(j.start,cb.BLOCKON))
negative_intron_data.append(TuringCode(j.stop,cb.BLOCKOFF))
cutoff=args.cutoff
coverage=0.0
for i,x in enumerate(codesToBedGraph(positive_data.iter())):
if x[SCORE_INDEX] >= cutoff:
retv.append((x[0],x[1],x[2],POSITIVE_STRAND,EXON_GROUP_CODE))
coverage+=float(x[1]-x[0])*x[2]/1000.0
for i,x in enumerate(codesToBedGraph(negative_data.iter())):
if x[SCORE_INDEX] >= cutoff:
retv.append((x[0],x[1],x[2],NEGATIVE_STRAND,EXON_GROUP_CODE))
coverage+=float(x[1]-x[0])*x[2]/1000.0
INTRON_CUTOFF=1.0
for i,x in enumerate(codesToBedGraph(positive_intron_data.iter())):
if x[SCORE_INDEX] >= INTRON_CUTOFF:
retv.append((x[0],x[1],x[2],POSITIVE_STRAND,INTRON_GROUP_CODE))
for i,x in enumerate(codesToBedGraph(negative_intron_data.iter())):
if x[SCORE_INDEX] >= INTRON_CUTOFF:
retv.append((x[0],x[1],x[2],NEGATIVE_STRAND,INTRON_GROUP_CODE))
retv.sort(key=itemgetter(0,1,2))
#TODO how to sort!
local_dbi.close()
return coverage,retv
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):
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(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 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 call_peaks(chrom,bedgraph,exon_cutoff):
#TODO
if hasGenome:
genome=DBI.init(args.genome,"genome")
def filter_low_complexity(peak):
new_exon_starts=[]
new_exon_sizes=[]
local_peak=list(peak)
offset=local_peak[START_INDEX]
for exon_start,exon_size in itertools.izip(local_peak[EXONSTARTS_INDEX],local_peak[EXONSIZES_INDEX]):
exon_s=offset+exon_start
exon_e=offset+exon_start+exon_size
seq=genome.query(BED3(chrom,exon_s,exon_e),method="seq")
if len(seq) > 100:
cplx=float(complexity(seq[0:100]))/100
else:
cplx=float(complexity(seq))/len(seq)
if cplx > CPLX_CUTOFF:
new_exon_starts.append(exon_start)
new_exon_sizes.append(exon_size)
if len(new_exon_sizes) > 0:
local_peak[EXONSIZES_INDEX]=tuple(new_exon_sizes)
local_peak[START_INDEX]=new_exon_starts[0]+offset
local_peak[STOP_INDEX]=new_exon_starts[-1]+offset+new_exon_sizes[-1]
shift_start=new_exon_starts[0]
for i0 in range(len(new_exon_starts)):
new_exon_starts[i0]-=shift_start
local_peak[EXONSTARTS_INDEX]=tuple(new_exon_starts)
return tuple(local_peak)
else:
return None
gap=10
pos_beds=[]
neg_beds=[]
peaks=[]
i_p=0
i_n=0
last_pos_stop=0
last_neg_stop=0
for i in bedgraph:
if i[STRAND_INDEX]==POSITIVE_STRAND:
if i[GROUP_INDEX]==EXON_GROUP_CODE:
if i[SCORE_INDEX] >= exon_cutoff:
if len(pos_beds)>0:
if i[START_INDEX]-last_pos_stop < gap or last_pos_stop==0:
pos_beds.append(i)
if last_pos_stop < i[STOP_INDEX]:
last_pos_stop=i[STOP_INDEX]
else:
peak=bedsToPeak(pos_beds,"p_"+str(i_p))
if peak is not None:
if hasGenome:
peak=filter_low_complexity(peak)
if peak is not None:
peaks.append(peak)
i_p+=1
else:
peaks.append(peak)
i_p+=1
pos_beds=[i]
last_pos_stop=i[STOP_INDEX]
else:
last_pos_stop=i[STOP_INDEX]
pos_beds.append(i)
elif i[GROUP_INDEX]==INTRON_GROUP_CODE:
if last_pos_stop < i[STOP_INDEX]:
last_pos_stop=i[STOP_INDEX]
pos_beds.append(i)
else:
if i[GROUP_INDEX]==EXON_GROUP_CODE:
if i[SCORE_INDEX] >= exon_cutoff:
if len(neg_beds)>0:
if i[START_INDEX]-last_neg_stop < gap or last_neg_stop==0:
neg_beds.append(i)
if last_neg_stop < i[STOP_INDEX]:
last_neg_stop=i[STOP_INDEX]
else:
peak=bedsToPeak(neg_beds,"n_"+str(i_n))
if peak is not None:
if hasGenome:
peak=filter_low_complexity(peak)
if peak is not None:
peaks.append(peak)
i_n+=1
else:
peaks.append(peak)
i_n+=1
neg_beds=[i]
last_neg_stop=i[STOP_INDEX]
else:
neg_beds.append(i)
last_neg_stop=i[STOP_INDEX]
elif i[GROUP_INDEX]==INTRON_GROUP_CODE:
if last_neg_stop < i[STOP_INDEX]:
last_neg_stop=i[STOP_INDEX]
neg_beds.append(i)
if len(pos_beds)>0:
peak=bedsToPeak(pos_beds,"p_"+str(i_p))
if peak is not None:
if hasGenome:
peak=filter_low_complexity(peak)
if peak is not None:
peaks.append(peak)
else:
peaks.append(peak)
if len(neg_beds)>0:
peak=bedsToPeak(neg_beds,"n_"+str(i_n))
if peak is not None:
if hasGenome:
peak=filter_low_complexity(peak)
if peak is not None:
peaks.append(peak)
else:
peaks.append(peak)
peaks.sort()
return peaks