def build_fwtrack(self, fhd):
"""Build FWTrackI from all lines, return a FWTrackI object.
Note: All ranges will be merged (exclude the same
range) then sorted after the track is built.
If both_strand is True, it will store strand information in
FWTrackI object.
if do_merge is False, it will not merge the same range after
the track is built.
"""
fwtrack = FWTrackI()
i = 0
m = 0
for thisline in fhd:
(chromosome, fpos, strand) = self.__fw_parse_line(thisline)
i += 1
if i == 1000000:
m += 1
logging.info(" %d" % (m * 1000000))
i = 0
if fpos == None or not chromosome:
continue
fwtrack.add_loc(chromosome, fpos, strand)
return fwtrack
def build_fwtrack(self, lfhd, rfhd, dist=200):
"""Build FWTrackI from all lines, return a FWTrackI object.
lfhd: the filehandler for left tag file
rfhd: the filehandler for right tag file
dist: the best distance between two tags in a pair
The score system for pairing two tags:
score = abs(abs(rtag-ltag)-200)+error4lefttag+error4righttag
the smaller score the better pairing. If the score for a
pairing is bigger than 200, this pairing will be discarded.
Note only the best pair is kept. If there are over two best
pairings, this pair of left and right tags will be discarded.
Note, the orders in left tag file and right tag file must
match, i.e., the Nth left tag must has the same name as the
Nth right tag.
Note, remove comment lines beforehand.
"""
fwtrack = FWTrackI()
i = 0
m = 0
lnext = lfhd.next
rnext = rfhd.next
self.dist = dist
try:
while 1:
lline = lnext()
rline = rnext()
(chromosome, fpos, strand) = self.__fw_parse_line(lline, rline)
i += 1
if i == 1000000:
m += 1
logging.info(" %d" % (m * 1000000))
i = 0
if not fpos or not chromosome:
continue
fwtrack.add_loc(chromosome, fpos, strand)
except StopIteration:
pass
return fwtrack
def build_fwtrack(self, fhd):
"""Build FWTrackI from all lines, return a FWTrackI object.
Note only the unique match for a tag is kept.
"""
fwtrack = FWTrackI()
i = 0
m = 0
for thisline in fhd:
(chromosome, fpos, strand) = self.__fw_parse_line(thisline)
i += 1
if i == 1000000:
m += 1
logging.info(" %d" % (m * 1000000))
i = 0
if not fpos or not chromosome:
continue
fwtrack.add_loc(chromosome, fpos, strand)
return fwtrack
def read_motif2 (motif_fhd,species,cutoff=0):
"""Read motif scan result, and return a WigTrackI object
containing the motif locations.
* If the motif scan data file is not big, use this function to
load the whole file into memory. It may be faster than
read_motif().
motif_fhd : a file handler for binary motif scan result
species : must be "mm8" for mouse or "hg18" for human
cutoff : cutoff for the motif scan score
"""
motif_range_list = FWTrackI(fw=0)
if species == "hg18":
chromosomes_fp = { # store start and number of file-pos for every chromosome in bin file
"chr1":[0,0],"chr2":[0,0],"chr3":[0,0],
"chr4":[0,0],"chr5":[0,0],"chr6":[0,0],
"chr7":[0,0],"chr8":[0,0],"chr9":[0,0],
"chr10":[0,0],"chr11":[0,0],"chr12":[0,0],
"chr13":[0,0],"chr14":[0,0],"chr15":[0,0],
"chr16":[0,0],"chr17":[0,0],"chr18":[0,0],
"chr19":[0,0],"chr20":[0,0],"chr21":[0,0],
"chr22":[0,0],"chrX":[0,0],"chrY":[0,0]
}
chromosomes = ["chr1","chr2","chr3","chr4","chr5","chr6",
"chr7","chr8","chr9","chr10","chr11","chr12",
"chr13","chr14","chr15","chr16","chr17","chr18",
"chr19","chr20","chr21","chr22","chrX","chrY"]
elif species == "mm8":
chromosomes_fp = { # store start and number of file-pos for every chromosome in bin file
"chr1":[0,0],"chr2":[0,0],"chr3":[0,0],
"chr4":[0,0],"chr5":[0,0],"chr6":[0,0],
"chr7":[0,0],"chr8":[0,0],"chr9":[0,0],
"chr10":[0,0],"chr11":[0,0],"chr12":[0,0],
"chr13":[0,0],"chr14":[0,0],"chr15":[0,0],
"chr16":[0,0],"chr17":[0,0],"chr18":[0,0],
"chr19":[0,0],"chrX":[0,0],"chrY":[0,0]
}
chromosomes = ["chr1","chr2","chr3","chr4","chr5","chr6",
"chr7","chr8","chr9","chr10","chr11","chr12",
"chr13","chr14","chr15","chr16","chr17","chr18",
"chr19","chrX","chrY"]
else:
raise Exception("Only hg18/mm8 supported!")
motif_fhd.seek(0)
data = motif_fhd.read()
# unpack the start pos
p = 0
for chromosome in chromosomes:
chromosomes_fp[chromosome][0] = upk("<i",data[p:p+4])[0]
p += 128
# calculate number of hits
total_motif_hits = 0
for i in range(len(chromosomes)-1):
mh = (chromosomes_fp[chromosomes[i+1]][0]-chromosomes_fp[chromosomes[i]][0])/8
chromosomes_fp[chromosomes[i]][1] = mh
total_motif_hits += mh
# last one
mh = (len(data)-chromosomes_fp[chromosomes[-1]][0])/8
chromosomes_fp[chromosomes[-1]][1]=mh
total_motif_hits += mh
# read and write
read_motif_hits = 0
portion = 0
p = 0
n=0
for chromosome in chromosomes:
p = chromosomes_fp[chromosome][0]
for i in range(chromosomes_fp[chromosome][1]):
read_motif_hits += 1
portion = float(read_motif_hits)/total_motif_hits
if LOG:
sys.stdout.write("\r %.1f%% %s" % (portion*100,"#"*int(portion*50)))
sys.stdout.flush()
loc = upk("<i",data[p:p+4])[0]
score = upk("<f",data[p+4:p+8])[0]
p += 8
if score < 0:
strand = 1
score = score*-1
else:
strand = 0
#ofhd.write("%s\t%d\t%d\t%s_%s_%d\t%.2f\t%s\n" % (chromosome,loc-1,loc+motif_len-1,motif,chromosome,i,score,strand))
if score > cutoff:
#print score,cutoff
n+=1
motif_range_list.add_loc(chromosome,loc-1,strand)
#print loc-1
if LOG : sys.stdout.write("\n")
data = None
motif_range_list.merge_overlap()
return motif_range_list