7
|
1 '''
|
|
2 find overlap and test signifiance
|
|
3 '''
|
|
4
|
|
5 import os,sys
|
|
6
|
|
7 def lineCount(filename):
|
|
8 if os.stat(filename).st_size == 0:
|
|
9 return 0
|
|
10 with open(filename) as f:
|
|
11 for i, l in enumerate(f):
|
|
12 pass
|
|
13 print i
|
|
14 return i+1
|
|
15
|
|
16 def intersect(fileA,fileB,outfile,fraction,reciprocal):
|
|
17 # return fileA intervals that overlap with interval in fileB
|
|
18 cmd = 'intersectBed -a '+fileA+' -b '+fileB + ' -u -wa -f '+fraction +' '+ reciprocal + '>'+outfile
|
|
19 #print cmd
|
|
20 os.system(cmd)
|
|
21
|
|
22 def shuffle(fileA,fileB,genomefile,fraction,reciprocal,N):
|
|
23 # shuffle fileA N times, return the distribution of overlaps
|
|
24 nOverlap = []
|
|
25 for i in range(N):
|
|
26 # shuffle fileA using shuffleBed
|
|
27 #cmd = 'shuffleBed -i '+fileA+' -g '+genomefile +'>fileA.shuffled'
|
|
28 # using random_interval.py
|
|
29 cmd = 'python /Users/xuebing/galaxy-dist/tools/mytools/random_interval.py '+fileA+' fileA.shuffled across '+genomefile
|
|
30 os.system(cmd)
|
|
31 intersect('fileA.shuffled',fileB,'tmp',fraction,reciprocal)
|
|
32 nOverlap.append(lineCount('tmp'))
|
|
33 os.system('rm tmp')
|
|
34 os.system('rm fileA.shuffled')
|
|
35 return nOverlap
|
|
36
|
|
37 def main():
|
|
38 fileA = sys.argv[1]
|
|
39 fileB = sys.argv[2]
|
|
40 outfile = sys.argv[3]
|
|
41 outplot = sys.argv[4]
|
|
42 outshuffle = sys.argv[5]
|
|
43 N = int(sys.argv[6]) # times to shuffle
|
|
44 genomefile = sys.argv[7]
|
|
45 fraction = sys.argv[8]
|
|
46 if len(sys.argv) == 10:
|
|
47 reciprocal = sys.argv[9] # can only be '-r'
|
|
48 else:
|
|
49 reciprocal = ''
|
|
50
|
|
51 #print sys.argv
|
|
52
|
|
53 # number of lines in input
|
|
54 nA = lineCount(fileA)
|
|
55 nB = lineCount(fileB)
|
|
56
|
|
57 # intersect on real data
|
|
58 intersect(fileA,fileB,outfile,fraction,reciprocal)
|
|
59 # number of overlaps
|
|
60 nOverlapReal = lineCount(outfile)
|
|
61
|
|
62 #print 'number of intervals in inputA that overlap with intervals in inputB:',nOverlapReal
|
|
63
|
|
64 # shuffle fileA to estimate background
|
|
65 nOverlapNull = shuffle(fileA,fileB,genomefile,fraction,reciprocal,N)
|
|
66 out = open(outshuffle,'w')
|
|
67 out.write("\t".join(map(str,nOverlapNull)))
|
|
68 out.close()
|
|
69
|
|
70 # plot histogram
|
|
71 rscript = open('tmp.r','w')
|
|
72 rscript.write("options(warn=-1)\n")
|
|
73 rscript.write("x0 <- "+str(nOverlapReal)+"\n")
|
|
74 rscript.write("x <- c("+','.join(map(str,nOverlapNull))+")\n")
|
|
75 rscript.write("library(MASS)\n")
|
|
76 rscript.write("pv <- min((1+sum(x>=x0))/length(x),(1+sum(x<=x0))/length(x))\n")
|
|
77 rscript.write("title <- paste('actual:chance = ',x0,':',format(mean(x),digits=1,nsmall=1),' = ',format(x0/mean(x),digits=1,nsmall=2),', p-value < ',pv,sep='')\n")
|
|
78 rscript.write("pdf('"+outplot+"')\n")
|
|
79 rscript.write("library(grid)\n")
|
|
80 rscript.write("library(VennDiagram)\n")
|
|
81 rscript.write("venn <- venn.diagram(x=list(A=1:"+str(nA)+",B="+str(nA-nOverlapReal+1)+":"+str(nA+nB-nOverlapReal)+"),filename=NULL,fill=c('red','blue'),col='transparent',alpha=0.5,label.col='black',cex=3,lwd=0,fontfamily='serif',fontface='bold',cat.col = c('red', 'blue'),cat.cex=3,cat.fontfamily='serif',cat.fontface='bold')\n")
|
|
82 rscript.write("grid.draw(venn)\n")
|
|
83 rscript.write("h <- hist(x,breaks=50,xlab='number of overlaps',ylab='frequency',main=title)\n")
|
|
84 rscript.write("plot(h$mids,h$counts,type='h',xlim=c(min(h$mids,x0),max(x0,h$mids)),ylim=c(0,max(h$counts)),xlab='number of overlaps',ylab='frequency',main=title)\n")
|
|
85 rscript.write("points(x0,0,col='red')\n")
|
|
86 rscript.write("dev.off()\n")
|
|
87 rscript.close()
|
|
88 os.system("R --vanilla < tmp.r")
|
|
89 os.system('rm tmp.r')
|
|
90 main()
|