def plot(pdb,):
import sys
sys.path.append('/home/people/tc/svn/tc_sandbox/misc/')
import gnuplot
prefix = pdb
gnuplot.scatter_plot_2d(
prefix,regression=True,xlabel='mode',ylabel='correlation',ymin=-1,ymax=1,
)
return
def plot(pdb, ):
import sys
sys.path.append('/home/people/tc/svn/tc_sandbox/misc/')
import gnuplot
prefix = pdb
gnuplot.scatter_plot_2d(
prefix,
regression=True,
xlabel='mode',
ylabel='correlation',
ymin=-1,
ymax=1,
)
return
def main():
for chromosome in range(1,22+1,)+['X','Y',]:
fp_in = 'out_GATK/sep/ApplyRecalibration.recalibrated.filtered.%s.vcf' %(chromosome)
fp_out = 'out_VCFtools/freq%s' %(chromosome)
if os.path.isfile('%s.frq' %(fp_out)):
continue
s = 'bsub \
-M4000000 -R\'select[mem>4000] rusage[mem=4000]\' \
vcftools \
--vcf %s \
--freq \
--out %s \
' %(fp_in, fp_out,)
os.system(s)
for chromosome in range(1,22+1,)+['X','Y',]:
fp_out = 'out_VCFtools/freq%s.frq' %(chromosome)
## if chromosome <= 11:
## continue
print fp_out
fd = open(fp_out,'r')
for line in fd: break
l_pos = []
l_MAF = []
for line in fd:
l = line.split()
pos = float(l[1])/10**6
MAF = l[-1][2:]
l_pos += [pos]
l_MAF += [MAF]
## lines += ['%s %s\n' %(pos,MAF,)]
## fd = open('gnuplot%s.data','w')
## fd.writelines(lines)
## fd.close()
gnuplot.scatter_plot_2d(
'MAF%s' %(chromosome),
l_pos, l_MAF,
xlabel = 'pos (Mbp)',
ylabel = 'MAF',
)
return
for i in range(len(l_overlaps)):
overlap = l_overlaps[i]
overlap_normalized = (overlap-min(l_overlaps))/(1.-min(l_overlaps))
min_dist = l_dist_min[i]
## if overlap_normalized < 0.5 and min_dist > 10:
## print pdb, overlap, overlap_normalized, min_dist
if overlap_normalized < 0.5 and min_dist > 30:
print pdb, overlap, overlap_normalized, min_dist
l_gnuplot += ['%s %s %s\n' %(min_dist,overlap_normalized,pdb,)]
print l_vicinal.count(True), len(l_vicinal), l_vicinal.count(True)/float(len(l_vicinal))
##print minmindist
prefix = 'gnuplot'
fd = open('%s.gnuplotdata' %(prefix),'w')
fd.writelines(l_gnuplot)
fd.close()
gnuplot.scatter_plot_2d(
prefix,
## bool_regression_linear = True,
xlabel='minimum distance to catalytic site residue(s)',
ylabel='overlap between apo and holo eigenvectors',
bool_remove = False,
)
print l_vicinal.count(True), len(l_vicinal), l_vicinal.count(True)/float(len(l_vicinal))
def plot(d_mmCIF_main,d_rmsds,):
l_pdbs = d_rmsds.keys()
l_pdbs.sort()
l_temperature = []
l_ph = []
l_resolution = []
d_spacegroup = {}
d_starting_model = {}
l_correl_T = [[],[],]
l_correl_pH = [[],[],]
l_correl_resol_max = [[],[],]
d_histo_pH = {}
d_histo_T = {}
d_histo_resol = {}
for i1 in range(len(l_pdbs)-1):
pdb1 = l_pdbs[i1]
spacegroup1 = core.parse_mmCIF_item(d_mmCIF_main[pdb1[:4]],'_symmetry.space_group_name_H-M',)
T1 = core.parse_mmCIF_item(d_mmCIF_main[pdb1[:4]],'_diffrn.ambient_temp',)
pH1 = core.parse_mmCIF_item(d_mmCIF_main[pdb1[:4]],'_exptl_crystal_grow.pH',)
starting_model1 = core.parse_mmCIF_item(d_mmCIF_main[pdb1[:4]],'_refine.pdbx_starting_model',)
resolution1 = core.parse_mmCIF_item(d_mmCIF_main[pdb1[:4]],'_refine.ls_d_res_high',)
for i2 in range(i1+1,len(l_pdbs)):
pdb2 = l_pdbs[i2]
spacegroup2 = core.parse_mmCIF_item(d_mmCIF_main[pdb2[:4]],'_symmetry.space_group_name_H-M',)
T2 = core.parse_mmCIF_item(d_mmCIF_main[pdb2[:4]],'_diffrn.ambient_temp',)
pH2 = core.parse_mmCIF_item(d_mmCIF_main[pdb2[:4]],'_exptl_crystal_grow.pH',)
starting_model2 = core.parse_mmCIF_item(d_mmCIF_main[pdb2[:4]],'_refine.pdbx_starting_model',)
resolution2 = core.parse_mmCIF_item(d_mmCIF_main[pdb2[:4]],'_refine.ls_d_res_high',)
rmsd = d_rmsds[pdb1][pdb2]
if rmsd > 1:
print pdb1, pdb2, rmsd
if T1 and T2:
T_diff = abs(float(T2)-float(T1))
l_temperature += ['%s %s\n' %(T_diff,rmsd),]
l_correl_T[0] += [T_diff]
l_correl_T[1] += [rmsd]
print T_diff, 10*round(T_diff/10.,0)
if not 10*round(T_diff/10.,0) in d_histo_T.keys():
d_histo_T[10*round(T_diff/10.,0)] = 0
d_histo_T[10*round(T_diff/10.,0)] += 1
if pH1 and pH2:
pH_diff = abs(float(pH2)-float(pH1))
l_ph += ['%s %s\n' %(pH_diff,rmsd),]
l_correl_pH[0] += [pH_diff]
l_correl_pH[1] += [rmsd]
if not pH_diff in d_histo_pH.keys():
d_histo_pH[pH_diff] = 0
d_histo_pH[pH_diff] += 1
resolution_max = max(resolution1,resolution2,)
l_resolution += ['%s %s\n' %(resolution_max,rmsd),]
if resolution_max != 'N/A':
l_correl_resol_max[0] += [float(resolution_max)]
l_correl_resol_max[1] += [rmsd]
if not round(float(resolution_max),0) in d_histo_resol.keys():
d_histo_resol[round(float(resolution_max),0)] = 0
d_histo_resol[round(float(resolution_max),0)] += 1
d_spacegroup = append_to_dictionary(d_spacegroup,spacegroup1,spacegroup2,rmsd,)
d_starting_model = append_to_dictionary(d_starting_model,starting_model1,starting_model2,rmsd,)
r1 = statistics.correlation(l_correl_T[0],l_correl_T[1],)
r2 = statistics.correlation(l_correl_pH[0],l_correl_pH[1],)
r3 = statistics.correlation(l_correl_resol_max[0],l_correl_resol_max[1],)
##
## plot histograms
##
for prefix,d in [
['deltapH',d_histo_pH,],
['deltaT',d_histo_T,],
['maxresolution',d_histo_resol,],
]:
l = []
l_diffs = d.keys()
l_diffs.sort()
for diff in l_diffs:
l += ['%s %s\n' %(diff,d[diff],)]
fd = open('histo_%s.txt' %(prefix),'w')
fd.writelines(l)
fd.close()
l = [
'set terminal postscript eps enhanced color "Helvetica"\n',
'set output "gnuplot.ps"\n',
'set size 3,3\n',
'set style data histogram\n',
'set xtics rotate\n',
'set xlabel "%s\n' %(prefix),
'set ylabel "count\n',
'plot "histo_%s.txt" u 2:xtic(1) t ""\n' %(prefix)
]
fd = open('tmp.txt','w')
fd.writelines(l)
fd.close()
os.system('gnuplot tmp.txt')
os.system('convert gnuplot.ps histo_%s.png' %(prefix))
##
## plot rmsd as a function of each property (2d)
##
for prefix,data,xlabel in [
['pH',l_ph,'pH diff',],
['Temperature',l_temperature,'T diff',],
['resolution',l_resolution,'maximum resolution',],
]:
prefix += method
fd = open('%s.gnuplotdata' %(prefix),'w')
fd.writelines(data)
fd.close()
gnuplot.scatter_plot_2d(
prefix,xlabel=xlabel,ylabel='RMSD %s' %(method,),
## averages=True,
regression=True,
)
##
## plot rmsd as a function of each property (contour)
##
for d,prefix in [
[d_spacegroup,'spacegroup',],
[d_starting_model,'startingmodel',],
]:
d_tics = {}
l_tics = d.keys()
l_tics.sort()
for i in range(len(l_tics)):
d_tics[l_tics[i]] = i+.5
z1 = 9
z2 = 0
l_data = []
for x in range(len(l_tics)):
k1 = l_tics[x]
for y in range(len(l_tics)):
k2 = l_tics[y]
if not k2 in d[k1].keys():
average = 9
else:
l_rmsds = d[k1][k2]
average = sum(l_rmsds)/len(l_rmsds)
if average < z1:
z1 = average
if average > z2:
z2 = average
l_data += ['%s %s %s\n' %(x,y,average,)]
l_data += ['%s %s %s\n' %(x,y+1,1,)]
l_data += ['\n']
for y in range(len(l_tics)):
l_data += ['%s %s %s\n' %(x+1,y,1,)]
l_data += ['%s %s %s\n' %(x+1,y+1,1,)]
l_data += ['\n']
gnuplot.contour_plot(
prefix,l_data,
title='%s %s' %(prefix,method,),zlabel='RMSD %s' %(method),
d_xtics = d_tics, d_ytics = d_tics,
palette = '0 1 0 0, 0.9999 0 0 1, 0.9999 1 1 1, 1 1 1 1',
z1 = z1, z2 = z2+0.1,
bool_remove = False,
)
os.system('convert %s.ps %s_spacegroup%s_mutations%s_atoms%s.png' %(prefix,prefix,spacegroup.replace(' ',''),n_mutations_max,method,))
## os.remove('%s.ps' %(prefix,))
print d_spacegroup
print d_starting_model
print r1
print r2
print r3
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
## print pdb, overlap, overlap_normalized, min_dist
if overlap_normalized < 0.5 and min_dist > 30:
print pdb, overlap, overlap_normalized, min_dist
l_gnuplot += ['%s %s %s\n' % (
min_dist,
overlap_normalized,
pdb,
)]
print l_vicinal.count(True), len(l_vicinal), l_vicinal.count(True) / float(
len(l_vicinal))
##print minmindist
prefix = 'gnuplot'
fd = open('%s.gnuplotdata' % (prefix), 'w')
fd.writelines(l_gnuplot)
fd.close()
gnuplot.scatter_plot_2d(
prefix,
## bool_regression_linear = True,
xlabel='minimum distance to catalytic site residue(s)',
ylabel='overlap between apo and holo eigenvectors',
bool_remove=False,
)
print l_vicinal.count(True), len(l_vicinal), l_vicinal.count(True) / float(
len(l_vicinal))
def plot_MDS(d_options):
l_cmds = []
bfile_in = d_options['bfile']
bfile_out = os.path.split(bfile_in)[1]
if not os.path.isfile('%s.mds' %(bfile_out)):
return
## count number of samples
n_samples = int(os.popen('cat %s.fam | wc -l' %(bfile_in)).read())
if d_options['remove'] != None:
execmd('cat %s | sort -k1,1 > remove.sorted' %(d_options['remove']))
execmd('cat %s.fam | sort -k1,1 > fam.sorted' %(d_options['bfile']))
execmd('join fam.sorted remove.sorted > %s.fam.joined' %(bfile_out))
n_samples -= int(os.popen('cat %s.fam.joined | wc -l' %(bfile_out)).read())
os.remove('remove.sorted')
os.remove('fam.sorted')
## sort
execmd('cat %s.mds | awk \'NR>1\' | sort -k1,1 > %s.mds.sorted' %(
bfile_out,bfile_out,))
execmd('sort -k1,1 samples2pops.dic > samples2pops.dic.sorted')
## join samples
cmd = 'join -a1 -e "Unknown" -o 1.1,1.2,1.3,1.4,1.5,1.6,1.7,2.2'
cmd += ' %s.mds.sorted samples2pops.dic.sorted' %(bfile_out)
cmd += ' > %s.mds.joined' %(bfile_out)
execmd(cmd)
if (
int(os.popen('cat %s.mds.joined | wc -l' %(bfile_out)).read())+1
!=
int(os.popen('cat %s.mds | wc -l' %(bfile_out)).read())
):
print int(os.popen('cat %s.mds.joined | wc -l' %(bfile_out)).read())+1
print int(os.popen('cat %s.mds | wc -l' %(bfile_out)).read())
stop
##
cmd = 'cat %s.mds.joined' %(bfile_out)
cmd += " | awk '{"
cmd += 'print $1,$2,$3,$4,$5,$6,$7 > $8".mds"'
cmd += "}'"
execmd(cmd)
l_pops = os.popen(
"cat %s.mds.joined | awk '{print $8}' | sort -u" %(bfile_out)
).read().strip().split('\n')
## define colors for set of populations
l_colors = [
[255,0,0,],
[255,85,0,],
[255,170,0,],
[255,255,0,],
[170,255,0,],
[85,255,0,],
[0,255,0,],
[0,255,85,],
[0,255,170,],
[0,255,255,],
[0,170,255,],
[0,85,255,],
[0,0,255,],
[85,0,255,],
[170,0,255,],
[255,0,255,],
[255,0,170,],
[255,0,85,],
[0,0,0,],
[85,85,85,],
[170,170,170,],
]
l_pts = [5,7,9,11]
line_plot = 'set key out\n'
line_plot += 'plot '
for i in xrange(len(l_pops)):
pop = l_pops[i]
color = "".join(map(chr, l_colors[i%len(l_colors)])).encode('hex')
pt = l_pts[i%len(l_pts)]
line_plot += '"%s.mds" u 4:5 pt %i ps 2 lc rgb "#%s" t "%s", ' %(
pop,pt,color,pop)
line_plot = line_plot[:-2]
c1 = 1
c2 = 2
gnuplot.scatter_plot_2d(
'%s.mds' %(bfile_out),
line_plot = line_plot,
xlabel = 'C%i' %(c1),
ylabel = 'C%i' %(c2),
title='%s' %(bfile_out),
prefix_out='mds.2D.%s.%i.%i' %(bfile_out,c1,c2),
bool_execute = False,
bool_remove = False,
)
return
def main(self,):
instance_GATK = GATK_pipeline.main()
d_chromosome_lengths = instance_GATK.parse_chromosome_ranges()
l_fn = os.listdir('stdout')
## for fn in l_fn:
## print fn
## if (
## fn[:len('UnifiedGenotyper')] == 'UnifiedGenotyper'
## and
## fn[len('UnifiedGenotyper')] != '.'
## ):
## old = os.path.join('stdout',fn)
## new = os.path.join('stdout',fn.replace('UnifiedGenotyper','UnifiedGenotyper.'))
## os.rename(old,new)
## if '99' in fn:
## fn_new = '.'.join([fn.split('.')[0],fn.split('.')[1],fn.split('.')[3],fn.split('.')[2],])
## old = os.path.join('stdout',fn)
## new = os.path.join('stdout',fn.replace(fn_new,''))
## print old
## print new
## stop
## os.rename(old,new)
## continue
## if (
## fn[:len('UnifiedGenotyper')] == 'UnifiedGenotyper'
## and
## fn[-1] != 't'
## ):
## old = os.path.join('stdout',fn)
## fn_new = '.'.join([fn.split('.')[0],fn.split('.')[1],fn.split('.')[3],fn.split('.')[2],])
## new = os.path.join('stdout',fn_new)
## os.rename(old,new)
## stop
d_resources = {'CPU':{},'Memory':{},}
l_fn.sort()
for fn in l_fn:
if os.path.isdir(os.path.join('stdout',fn)):
continue
## print fn
index1 = fn.index('.')
step = fn[:index1]
if '.' in fn[index1+1:]:
index2 = index1+fn[index1+1:].index('.')+1
chromosome = fn[index1+1:index2]
else:
chromosome = ''
if chromosome == '23': chromosome = 'X'
if chromosome == '24': chromosome = 'Y'
fd = open('stdout/%s' %(fn),'r')
lines = fd.readlines()
fd.close()
## it would be faster to do rindex instead of regex,
## but in a few cases rubbish was appended to the farm log files
keyword1 = re.compile(r' Max Memory :')
keyword2 = re.compile(r' CPU time :')
l_mem = []
l_cpu = []
for line in lines:
result1 = keyword1.search(line)
result2 = keyword2.search(line)
if result1 or result2:
v = float(line.split(':')[1].replace('sec.','').replace('MB',''))
if result1: l_mem += [v]
else: l_cpu += [v]
cpu = max(l_cpu)
mem = l_mem[l_cpu.index(cpu)]
## if 'ApplyRecalibration' in fn and cpu > 1:
if 'VariantRecalibrator' in fn and cpu > 1:
print '%4i %4i %s' %(int(mem), int(cpu), chromosome), fn
## ignore if took less than a minute
if cpu < 60:
if os.path.getsize(os.path.join('stdout',fn)) < 2200:
print os.path.getsize(os.path.join('stdout',fn)), fn
stop
os.remove(os.path.join('stdout',fn))
continue
continue
for k_resource, v_resource in [
['CPU',cpu,],
['Memory',mem,],
]:
if not step in d_resources[k_resource].keys():
d_resources[k_resource][step] = {}
if not chromosome in d_resources[k_resource][step].keys():
d_resources[k_resource][step][chromosome] = []
elif step not in ['UnifiedGenotyper','IMPUTE2',]:
print step, chromosome, k_resource, v_resource
d_resources[k_resource][step][chromosome] += [v_resource]
for k_resource in d_resources.keys():
for step in d_resources[k_resource].keys():
if 'Downsample' in step: continue
if 'samtools' in step: continue
l_y = []
for chromosome in d_resources[k_resource][step].keys():
y = usage = d_resources[k_resource][step][chromosome]
if k_resource == 'CPU':
y = sum(y)/3600.
elif k_resource == 'Memory':
y = (sum(y)/len(y))
else:
print k_resource
stop
## lines += ['%s %s\n' %(x,y,)]
l_y += [y]
if k_resource == 'CPU':
print k_resource, step, sum(l_y), len(l_y)
else:
print k_resource, step, sum(l_y)/len(l_y), len(l_y)
d_labels = {'Memory':'Mb','CPU':'hours'}
for k_resource in d_resources.keys():
for step in d_resources[k_resource].keys():
l_x = []
l_y = []
if len(d_resources[k_resource][step].keys()) <= 3:
continue
for chromosome in d_resources[k_resource][step].keys():
if chromosome == '':
continue
if chromosome[0] == '_':
continue
x = chromosome_length = d_chromosome_lengths[chromosome]/(10**6)
y = usage = d_resources[k_resource][step][chromosome]
if k_resource == 'CPU':
y = sum(y)/3600.
elif k_resource == 'Memory':
y = (sum(y)/len(y))
else:
print k_resource
stop
## lines += ['%s %s\n' %(x,y,)]
l_x += [x]
l_y += [y]
if len(l_x) == 0:
print step
continue
## fd = open('gnuplot_%s_%s.data' %(k_resource,step,),'w')
## fd.writelines(lines)
## fd.close()
prefix = '%s_%s' %(k_resource,step,)
print 'plotting', k_resource, step
if k_resource == 'CPU':
print k_resource, step, sum(l_y), len(l_y)
else:
print k_resource, step, sum(l_y)/len(l_y), len(l_y)
gnuplot.scatter_plot_2d(
prefix,l1=l_x,l2=l_y,
ylabel='%s (%s)' %(k_resource,d_labels[k_resource]),
xlabel='chromosome length (Mbp)',
title=prefix.replace('_',' '),
)
return
gnuplot.scatter_plot_2d(
prefix,
xlabel = 'RMSD_C_{/Symbol a} / @^{/Symbol \ \260}A', ## {\305} is Angstrom if iso encoding
ylabel = '<{/Symbol Dc_1}> / {/Symbol \260}',
xmin = 0, ymin = 0,
xmax = xmax,
ymax = ymax,
## ylabel = 'heavy atom RMSD',
## ymax = ymax,
## function = 'x',
key_vert_pos = 'left',
## title = 'alpha carbon and heavy atom RMSD between %s wt structures' %(protein),
title = title,
bool_multiple_columns = True,
## d_columns = d_columns,
l_columns = l_columns,
l_colors = l_colors,
l_pointtypes = l_pointtypes,
l_pointsizes = l_pointsizes,
pointsize = 1,
## bool_title = False,
bool_remove = False,
)
def MDS():
import gnuplot
prefix = 'Baganda29_quad29_octo200'
prefix = 'Baganda29_quad29_octo200_excldiscordant'
prefix = 'Baganda29_quad29_octo200_SNPQCtogether'
## prefix = 'Ga-Adangbe'
## prefix = 'Zulu'
## prefix = 'Ga-Adangbeexcldiscordant'
## prefix = 'Zuluexcldiscordant'
bool_exclude_discordant = False
if 'excldiscordant' in prefix:
bool_exclude_discordant = True
bool_merge = True
if prefix in [
'Baganda29_quad29_octo200_SNPQCtogether',
'Ga-Adangbe','Zulu',
'Ga-Adangbeexcldiscordant','Zuluexcldiscordant',
]:
bool_merge = False
if prefix == 'Baganda29_quad29_octo200_SNPQCtogether':
bfile = 'pops/Baganda_quad29octo200/Baganda_quad29octo200.SNPQC'
else:
if 'excldiscordant' in prefix:
bfile = 'pops/%s/%s.SNPQC' %(
prefix.replace('excldiscordant',''),
prefix.replace('excldiscordant',''),
)
else:
bfile = 'pops/%s/%s.SNPQC' %(prefix,prefix,)
fn_ld_regions = 'pops/Baganda_octo/Baganda_octo.ldregions.SNPs'
##
## find common SNPs post QC
##
if not os.path.isfile('%s.extract' %(prefix)):
if bool_merge == False:
cmd = 'cat %s.bim > %s.extract' %(bfile,prefix)
execmd(cmd)
else:
cmd = "cat pops/Baganda_quad/Baganda_quad.SNPQC.bim | awk '{print $2}' | sort > Baganda_quad.SNPs"
execmd(cmd)
cmd = "cat pops/Baganda_octo/Baganda_octo.SNPQC.bim | awk '{print $2}' | sort > Baganda_octo.SNPs"
execmd(cmd)
cmd = 'comm -12 Baganda_quad.SNPs Baganda_octo.SNPs > %s.extract' %(prefix)
execmd(cmd)
os.remove('Baganda_quad.SNPs')
os.remove('Baganda_octo.SNPs')
##
## find common samples post QC
##
cmd = 'cat pops/Baganda_quad/Baganda_quad.SNPQC.fam | sort > Baganda_quad.fam'
execmd(cmd)
cmd = 'cat pops/Baganda_octo/Baganda_octo.SNPQC.fam | sort > Baganda_octo.fam'
execmd(cmd)
cmd = "cat Baganda_quad.fam | awk '{print substr($1,12,10)}' | sort > Baganda_quad.samples"
execmd(cmd)
cmd = "cat Baganda_octo.fam | awk '{print substr($1,12,10)}' | sort > Baganda_octo.samples"
execmd(cmd)
cmd = 'comm -12 Baganda_quad.samples Baganda_octo.samples > Baganda29.samples'
execmd(cmd)
os.remove('Baganda_quad.samples')
os.remove('Baganda_octo.samples')
cmd = 'fgrep -f Baganda29.samples Baganda_quad.fam | sort > Baganda29_quad29_octo0.fam'
execmd(cmd)
cmd = 'fgrep -f Baganda29.samples Baganda_octo.fam | sort > Baganda29_quad0_octo29.fam'
execmd(cmd)
os.remove('Baganda29.samples')
cmd = 'comm -23 Baganda_quad.fam Baganda29_quad29_octo0.fam > Baganda29_quad71_octo0.fam'
execmd(cmd)
cmd = 'comm -23 Baganda_octo.fam Baganda29_quad0_octo29.fam > Baganda29_quad0_octo200.fam'
execmd(cmd)
cmd = 'cat Baganda29_quad0_octo200.fam Baganda29_quad29_octo0.fam > Baganda29_quad29_octo200.fam'
execmd(cmd)
cmd = 'cat Baganda29_quad0_octo200.fam Baganda29_quad29_octo0.fam > Baganda29_quad29_octo200_excldiscordant.fam'
execmd(cmd)
if bool_merge == False:
cmd = 'cat %s.fam > %s.fam' %(bfile,prefix,)
execmd(cmd)
##
## --bmerge
##
if not os.path.isfile('%s.bed' %(prefix)):
cmd = 'plink \\\n'
if bool_merge == False:
cmd += '--bfile %s \\\n' %(bfile)
else:
cmd += '--bfile pops/Baganda_quad/Baganda_quad.SNPQC \\\n'
cmd += '--bmerge \\\n'
cmd += 'pops/Baganda_octo/Baganda_octo.SNPQC.bed \\\n'
cmd += 'pops/Baganda_octo/Baganda_octo.SNPQC.bim \\\n'
cmd += 'pops/Baganda_octo/Baganda_octo.SNPQC.fam \\\n'
cmd += '--keep %s.fam \\\n' %(prefix)
cmd += '--extract %s.extract \\\n' %(prefix)
## cmd += '--exclude 26diff_and_monomorphic.SNPs \\\n' ## tmp
if bool_exclude_discordant == True:
cmd += '--exclude discordant.SNPs \\\n' ## tmp
cmd += '--make-bed --out %s \\\n' %(prefix)
execmd(cmd)
##
## --indep-pairwise
##
if not os.path.isfile('%s.prune.in' %(prefix)):
cmd = 'plink \\\n'
cmd += '--bfile %s \\\n' %(prefix)
cmd += '--out %s \\\n' %(prefix)
## settings
cmd += '--indep-pairwise 50 5 0.2 \\\n'
cmd += '--maf 0.05 \\\n'
## SNP exclusion
cmd += '--exclude %s \\\n' %(fn_ld_regions)
execmd(cmd)
##
## --genome
##
if not os.path.isfile('%s.genome' %(prefix)):
cmd = 'plink \\\n'
cmd += '--bfile %s \\\n' %(prefix)
cmd += '--out %s \\\n' %(prefix)
cmd += '--genome \\\n'
## SNP exclusion
cmd += '--extract %s.prune.in \\\n' %(prefix)
cmd += '--exclude %s \\\n' %(fn_ld_regions)
execmd(cmd)
##
## --cluster
##
if not os.path.isfile('%s.mds' %(prefix)):
cmd = 'plink \\\n'
cmd += '--bfile %s \\\n' %(prefix)
cmd += '--out %s \\\n' %(prefix)
cmd += '--cluster \\\n'
cmd += '--mds-plot 4 \\\n'
cmd += '--read-genome %s.genome \\\n' %(prefix)
## SNP exclusion
cmd += '--extract %s.prune.in \\\n' %(prefix)
cmd += '--exclude %s \\\n' %(fn_ld_regions)
execmd(cmd)
## ##
## ## EIGENSOFT
## ##
## eigensoft(prefix,fn_ld_regions,)
##
## plot
##
if not os.path.isfile('%s.mds' %(prefix)):
sys.exit(0)
if bool_merge == True:
cmd = "cat Baganda29_quad29_octo0.fam | awk '{print $1}' > Baganda29_quad29.samples"
execmd(cmd)
cmd = "cat Baganda29_quad0_octo29.fam | awk '{print $1}' > Baganda29_octo29.samples"
execmd(cmd)
for suffix in ['quad','octo',]:
cmd = 'fgrep -f Baganda29_%s29.samples %s.mds' %(suffix,prefix,)
cmd += " | awk '{print substr($1,12,10),$4,$5}'"
cmd += ' | sort -k1,1'
cmd += ' > %s_%s29.mds' %(prefix,suffix)
execmd(cmd)
if suffix == 'quad':
continue
cmd = "cat pops/Baganda_%s/Baganda_%s.fam | awk '{print $1}' > Baganda29_%s.samples" %(
suffix,suffix,suffix,)
execmd(cmd)
cmd = 'fgrep -f Baganda29_%s.samples %s.mds' %(suffix,prefix,)
cmd += " | awk '{print substr($1,12,10),$4,$5}'"
cmd += ' | sort -k1,1'
cmd += ' > %s_%s.mds' %(prefix,suffix)
execmd(cmd)
## lines_extra = ['set key out\n']
## fd = open('%s_quad29.mds' %(prefix))
## lines4 = fd.readlines()
## fd.close()
## fd = open('%s_octo29.mds' %(prefix))
## lines8 = fd.readlines()
## fd.close()
## for i in xrange(len(lines4)):
## l4 = lines4[i].split()
## l8 = lines8[i].split()
## x4 = float(l4[1])
## y4 = float(l4[2])
## x8 = float(l8[1])
## y8 = float(l8[2])
## lines_extra += ['set arrow from %f,%f to %f,%f\n' %(x4,y4,x8,y8,)]
n_samples = int(os.popen('cat %s.mds | wc -l' %(prefix)).read())-1
## without pruning
n_SNPs = int(os.popen('cat %s.bim | wc -l' %(prefix)).read())
## with pruning
n_SNPs = int(os.popen('cat %s.prune.in | wc -l' %(prefix)).read())
if bool_merge == False:
execmd("cat omni2.5-4_20120904_agv_gtu.fam | awk '{print $2}' > quad.samples")
execmd("cat omni2.5-8_agv_20120910_gtu.fam | awk '{print $2}' > octo.samples")
execmd('fgrep -w -f quad.samples %s.mds > %s_quad.mds' %(prefix,prefix,))
execmd('fgrep -w -f octo.samples %s.mds > %s_octo.mds' %(prefix,prefix,))
line_plot = 'plot '
line_plot += '"%s_quad.mds" u 4:5 ps 2 pt 7 lc 1 t "quad",' %(prefix)
line_plot += '"%s_octo.mds" u 4:5 ps 2 pt 7 lc rgb "#0000FF" t "octo",' %(prefix)
line_plot = line_plot[:-1]
else:
line_plot = 'plot '
line_plot += '"%s_quad29.mds" u 2:3 ps 2 pt 7 lc 1 t "quad",' %(prefix)
## line_plot += '"%s_octo29.mds" u 2:3 ps 3 pt 7 lc 3 t "octo",' %(prefix)
line_plot += '"%s_octo.mds" u 2:3 ps 2 pt 7 lc rgb "#0000FF" t "octo",' %(prefix)
line_plot = line_plot[:-1]
gnuplot.scatter_plot_2d(
'%s.mds' %(prefix),
line_plot = line_plot,
## column1 = 2, column2 = 3,
xlabel = 'C1',
ylabel = 'C2',
title='%s (n_{samples}=%i, n_{SNPs}=%i)' %(
'Baganda29',n_samples,n_SNPs,
),
prefix_out='%s.mds' %(prefix),
## lines_extra=lines_extra,
bool_remove=False,
)
return
if pdb in ['2d4i_b','2d4k_n',]:
res_no -= 200
res_symbol1 = mutation[-2]
res_symbol2 = mutation[-1]
if pdb in d_mutants.keys():
mutation = '%1s%i%1s' %(res_symbol2,res_no,res_symbol1,)
d_ddG[mutation]['backward'] += [ddG]
else:
mutation = '%1s%i%1s' %(res_symbol1,res_no,res_symbol2,)
d_ddG[mutation]['forward'] += [ddG]
l = []
for mutation in d_ddG.keys():
for ddG_forward in d_ddG[mutation]['forward']:
for ddG_backward in d_ddG[mutation]['backward']:
l += ['%s %s\n' %(-ddG_backward,ddG_forward,)]
fd = open('UFFBAPS.gnuplotdata','w')
fd.writelines(l)
fd.close()
prefix = 'UFFBAPS'
xlabel = 'ddG backward'
ylabel = 'ddG forward'
gnuplot.scatter_plot_2d(
prefix, xlabel=xlabel, ylabel=ylabel,
bool_multiple_columns = False,
function = 'x',
bool_remove = False,
)