#!/software/bin/python
## T. Carstensen (tc9), M.S. Sandhu (ms23), D. Gurdasani (dg11)
## Wellcome Trust Sanger Institute, 2012
import os, math
import sys
sys.path.append('/nfs/users/nfs_t/tc9/github/ms23/GATK_pipeline')
import GATK_pipeline
sys.path.append('/nfs/users/nfs_t/tc9/github/tc9/math')
import statistics
instance_statistics = statistics.tests()
def main():
instance = GATK_pipeline.main()
d_chromosome_lengths = instance.parse_chromosome_ranges()
d_centromere_ranges = parse_centromere_ranges()
ped = 'omni2.5-8_20120516_gwa_ugand_gtu_autosomes_postsampqc_postsnpqc_flipped'
sepjoin = 'sep'
transpose(ped)
IMPUTE2_tped(ped,d_chromosome_lengths,sepjoin,d_centromere_ranges,)
BEAGLE_tped(ped,d_chromosome_lengths,sepjoin,d_centromere_ranges,)
return
def singlevcf_vs_multiplevcfs():
'''count variant call differences between mpileup and unifiedgenotyper'''
fp2 = 'out_GATK/join/ApplyRecalibration.recalibrated.filtered.vcf'
fp1 = 'out_mp15/vqsr/$CHROMOSOME.vqsr.filt.vcf'
fd2 = open(fp2,'r')
##
## parse first line of second file
##
for line2 in fd2:
l2, CHROM2, POS2, REF2, ALT2, FILTER2, bool_continue = parse_line_vcf(line2)
if bool_continue == True:
continue
break
## count1 = 0
## count2 = 0
## count_intersect = 0
## count_unique1 = 0
## count_unique2 = 0
## for chromosome in [str(i) for i in range(1,22+1,)]+['X','Y',]:
## print chromosome
## l_pos2 = [POS2]
## for line2 in fd2:
## l2, CHROM2, POS2, REF2, ALT2, FILTER2, bool_continue = parse_line_vcf(line2)
## if bool_continue == True: continue
## if CHROM2 != chromosome: break
## l_pos2 += [POS2]
## l_pos1 = []
## fd1 = open('out_mp15/vqsr/%s.vqsr.filt.vcf' %(chromosome),'r')
## for line2 in fd1:
## l2, CHROM2, POS2, REF2, ALT2, FILTER2, bool_continue = parse_line_vcf(line2)
## if bool_continue == True: continue
## l_pos1 += [POS2]
## count1 += len(l_pos1)
## count2 += len(l_pos2)
## set1 = set(l_pos1)
## del l_pos1
## set2 = set(l_pos2)
## del l_pos2
## count_intersect += len(set1&set2)
## count_unique1 += len(set1-set2)
## count_unique2 += len(set2-set1)
## del set1
## del set2
## print count1
## print count2
## print count_intersect
## print count_unique1, count1-count_intersect
## print count_unique2, count2-count_intersect
## stop
d_QUAL = {}
count_intersect = 0
count1 = 0
count2 = 1
l_chromosomes = [str(i) for i in range(1,22+1,)]+['X','Y',]
for chromosome in l_chromosomes:
d_QUAL[chromosome] = []
fd1 = open('out_mp15/vqsr/%s.vqsr.filt.vcf' %(chromosome),'r')
for line1 in fd1:
l1, CHROM1, POS1, REF1, ALT1, FILTER1, bool_continue = parse_line_vcf(line1)
if bool_continue == True:
continue
count1 += 1
## if CHROM1 == '2' or CHROM2 == '2':
## break
## end of vcf2
if CHROM2 == None:
continue
if CHROM1 != CHROM2:
## loop over lines2
if l_chromosomes.index(CHROM1) > l_chromosomes.index(CHROM2):
for line2 in fd2:
l2, CHROM2, POS2, REF2, ALT2, FILTER2, bool_continue = parse_line_vcf(line2)
if bool_continue == True:
continue
count2 += 1
if CHROM1 != CHROM2:
continue
## bool_chromosome_diff = True
break
## loop over lines1
else:
## bool_chromosome_diff = True
continue
if POS1 == POS2:
if l1[5] != '999' and l2[5] != '999':
d_QUAL[CHROM1] += ['%s %s\n' %(l1[5],l2[5],)]
count_intersect += 1
for line2 in fd2:
l2, CHROM2, POS2, REF2, ALT2, FILTER2, bool_continue = parse_line_vcf(line2)
if bool_continue == True:
continue
count2 += 1
break
continue
## loop over lines1
elif POS2 > POS1:
if CHROM1 != CHROM2:
print 'b', CHROM1, CHROM2
stop
continue
## loop over lines2
## elif POS1 > POS2:
else:
for line2 in fd2:
l2, CHROM2, POS2, REF2, ALT2, FILTER2, bool_continue = parse_line_vcf(line2)
if bool_continue == True:
continue
count2 += 1
if CHROM1 != CHROM2:
print 'c', CHROM1, CHROM2
stop
else:
## loop over lines1
if POS2 > POS1:
break
elif POS1 == POS2:
if l1[5] != '999' and l2[5] != '999':
d_QUAL[CHROM1] += ['%s %s\n' %(l1[5],l2[5],)]
count_intersect += 1
for line2 in fd2:
l2, CHROM2, POS2, REF2, ALT2, FILTER2, bool_continue = parse_line_vcf(line2)
if bool_continue == True:
continue
count2 += 1
break
break
else:
continue
fd1.close()
fd2.close()
print count_intersect
print count1
print count2
fd = open('QUAL.gnuplotdata','w')
for chromosome in d_QUAL.keys():
print chromosome, len(d_QUAL[chromosome])
fd.writelines(d_QUAL[chromosome])
fd.close()
gnuplot.scatter_plot_2d(
'QUAL',regression=True,
xlabel='QUAL UnifiedGenotyper',
ylabel='QUAL mpileup',
)
l_QUAL1 = []
l_QUAL2 = []
for chromosome in d_QUAL.keys():
for line in d_QUAL[chromosome]:
l = line.split()
QUAL1 = float(l[0])
QUAL2 = float(l[1])
l_QUAL1 += [QUAL1]
l_QUAL2 += [QUAL2]
instance = statistics.tests()
r = instance.correlation(l_QUAL1,l_QUAL2,)
print r
stop
return