y[snp] = [genos[3], 'A', 'G'] + genos[6:] print >> currbimbam, ", ".join(y) #t1 = time.time() #print t1-t0 currbimbam.close() #genomat = matrix_reader(genodir + 'hutt.imputed.dhssnps.bimbam',sep=",") print "Running GEMMA..." gemmer = (hmdir + 'Programs/gemma0.94 -g ' + currfiles + '.bimbam -p ' + currfiles + '.pheno -k ' + currfiles + '.square.txt -c ' + currfiles + '.covariates -lmm 4 -maf 0.05 -o curr_' + pheno) t0 = time.time() ifier(gemmer) t1 = time.time() print t1-t0 #currresults = open(genodir + 'output/curr_' + pheno + '.assoc.txt','r') currresults = matrix_reader(genodir + 'output/curr_' + pheno + '.assoc.txt',dtype='f8') currsort = curresults[currresults[:,12].argsort()] currwins = currsort[0:100,] currscores = [0]*len(dhsdic[dhsdic.keys()[0]]) for snp in currwins[:,1]: currscores = currscores + dhsdic[snp] currperms = [0]*len(dhsdic[dhsdic.keys()[0]]) currpermwins = [0]*len(dhsdic[dhsdic.keys()[0]]) curractive = [0]*len(dhsdic[dhsdic.keys()[0]]) print "Running permutations..." for perm in xrange(0,100): tissueps = permer(scores = currscores, circuit = currperms, permwins = currpermwins, actives=curractive) #completedgenes += 1 #t1 = time.time()
genodir = sys.argv[2]
outname = sys.argv[3]
genor = genodir + 'ByChr/' + outname + '.chr' + chrm + '.genos.txt'
anoter = genodir + 'ByChr/' + outname + '.chr' + chrm + '.annot.txt'
print "Loading annotations..."
sys.stdout.flush()
bim = open(genodir + 'ByChr/' + outname + '.chr' + chrm + '.bim','r')
bimer = bim.readlines()
snpids = [x.strip().split()[1] for x in bimer]
snppos = [x.strip().split()[3] for x in bimer]
print "Loading genotypes..."
sys.stdout.flush()
raws = matrix_reader(genodir + 'ByChr/' + outname + '.chr' + chrm + '.raw',sep=" ")
#linecounter = subprocess.Popen('wc -l ' + genodir + 'ByChr/hutt.imputed.chr' + chrm + '.raw', shell=True, stdout=subprocess.PIPE)
#linecount = int(linecounter.communicate()[0].strip().split()[0])
#columncounter = subprocess.Popen('awk -F" " \'{print NF;exit}\' ' + genodir + 'ByChr/hutt.imputed.chr' + chrm + '.raw', shell=True, stdout=subprocess.PIPE)
#columncount = int(columncounter.communicate()[0].strip().split()[0])
#raws = numpy.zeros((linecount,columncount),dtype='|S2')
#rawin = open(genodir + 'ByChr/hutt.imputed.chr' + chrm + '.raw','r')
#for i,line in enumerate(rawin):
# raws[i,:] = line.strip().split()
#raws = numpy.loadtxt(genodir + 'ByChr/hutt.imputed.chr' + chrm + '.raw',dtype='str')
print "Transposing genotypes..."
sys.stdout.flush()
genorfile = open(genor,'w')
for line in range(6,raws.shape[1]):
import sys
import os
import numpy
sys.path.append('/mnt/lustre/home/cusanovich/Programs/')
sys.path.append('/mnt/lustre/home/cusanovich/Programs/lib/python2.6/site-packages/')
from DarrenTools import ifier, matrix_reader
genodir = '/mnt/lustre/home/cusanovich/500HT/Imputed1415/'
os.chdir(genodir)
#plinker = 'plink --bfile 500HT/Imputed1415/hutt.imputed.rename --recodeA'
#ifier(plinker)
genomatix = matrix_reader('/mnt/lustre/home/cusanovich/500HT/Imputed1415/hutt.imputed.rename.add.raw',sep=" ")
properg = genomatix.T[6:,1:]
snps = [x.split('_')[0] for x in genomatix.T[6:,0]]
findivs = genomatix.T[1,1:]
numpy.savetxt('/mnt/lustre/home/cusanovich/500HT/Imputed1415/hutt.imputed.rename.genos.bimbam', properg, fmt = '%s', delimiter = '\t', newline = '\n')
snper = open('/mnt/lustre/home/cusanovich/500HT/Imputed1415/hutt.imputed.rename.snps.bimbam','w')
print >> snper, '\n'.join(snps)
snper.close()
finer = open('/mnt/lustre/home/cusanovich/500HT/Imputed1415/hutt.imputed.rename.findivs.bimbam','w')
print >> finer, '\n'.join(findivs)
finer.close()
import os
import numpy
sys.path.append('/mnt/lustre/home/cusanovich/Programs/')
sys.path.append(
'/mnt/lustre/home/cusanovich/Programs/lib/python2.6/site-packages/')
from DarrenTools import ifier, matrix_reader
genodir = '/mnt/lustre/home/cusanovich/500HT/Imputed1415/'
os.chdir(genodir)
#plinker = 'plink --bfile 500HT/Imputed1415/hutt.imputed.rename --recodeA'
#ifier(plinker)
genomatix = matrix_reader(
'/mnt/lustre/home/cusanovich/500HT/Imputed1415/hutt.imputed.rename.add.raw',
sep=" ")
properg = genomatix.T[6:, 1:]
snps = [x.split('_')[0] for x in genomatix.T[6:, 0]]
findivs = genomatix.T[1, 1:]
numpy.savetxt(
'/mnt/lustre/home/cusanovich/500HT/Imputed1415/hutt.imputed.rename.genos.bimbam',
properg,
fmt='%s',
delimiter='\t',
newline='\n')
snper = open(
'/mnt/lustre/home/cusanovich/500HT/Imputed1415/hutt.imputed.rename.snps.bimbam',
'w')
gemmer = (hmdir + 'Programs/gemma0.94 -g ' + genodir + 'perm_curr_' + chrm + '_pc' + str(pcs) + '.bimbam -p ' + currfiles + '.pheno -k ' + currfiles + '.square.txt -c ' + currfiles + '.pcs.txt' + ' -lmm 4 -maf 0.05 -o perm_curr_' + chrm + '_pc' + str(pcs))
ifier(gemmer)
permering = open(genodir + 'output/perm_curr_' + chrm + '_pc' + str(pcs) + '.assoc.txt','r')
permers = [x.strip().split()[12] for x in permering.readlines()]
permers = [float(x) for x in permers if x != 'nan' and x != 'p_lrt']
permering.close()
permlow = min(permers)
if permlow <= pmin:
winnerperms += 1
if winnerperms == 10:
return 11/uniform(perm+2,perm+3)
return (winnerperms + 1)/float(10001)
print "Loading expression..."
if regressPCs:
mastercols = matrix_reader('/mnt/lustre/home/cusanovich/500HT/hutt' + mapper + '.' + distance + '.mastercols.txt',dtype='|S15')
if not regressPCs:
mastercols = matrix_reader('/mnt/lustre/home/cusanovich/500HT/hutt' + mapper + '.' + distance + '.chrmspecific.mastercols.txt',dtype='|S15')
masterdic = {}
exprcoldic = {}
chrmdic = {}
for i in range(mastercols.shape[0]):
try:
masterdic[mastercols[i,0]].append(mastercols[i,1])
except KeyError:
masterdic[mastercols[i,0]] = [mastercols[i,1]]
exprcoldic[mastercols[i,0]] = mastercols[i,2]
chrmdic[mastercols[i,0]] = mastercols[i,4]
glist = open(currfiles + '.gs','w')
for snp in masterdic.keys():
#for snp in masterdic.keys()[0:1000]:
print >> snplist, snp
print >> alist, 'A'
print >> glist, 'G'
snplist.close()
alist.close()
glist.close()
print "Running permutations..."
blocker = open(genodir + 'Block_' + blocknum + 'permwins.txt','w')
for perm in xrange(0,100):
permer(perm,randind = genoinds)
permresults = matrix_reader(genodir + 'output/perm_curr_' + blocknum + '.assoc.txt',dtype='f8')
permsort = permresults[permresults[:,12].argsort()]
permwins = permsort[0:100,]
print >> blocker, '\t'.join(permwins)
blocker.close()
cleanup = 'rm ' + genodir + '*curr_' + pheno + '.*'
ifier(cleanup)
print "Writing results..."
aller = open('/mnt/lustre/home/cusanovich/500HT/Tissues/' + pheno + '.enrichmentps.txt','w')
for i in xrange(0,len(tissueps)):
print >> aller, '{0}\t{2:.4g}'.format(dhsdic['rsID'][i],tissueps[i])
aller.close()
genodir = sys.argv[2]
outname = sys.argv[3]
genor = genodir + 'ByChr/' + outname + '.chr' + chrm + '.genos.txt'
anoter = genodir + 'ByChr/' + outname + '.chr' + chrm + '.annot.txt'
print "Loading annotations..."
sys.stdout.flush()
bim = open(genodir + 'ByChr/' + outname + '.chr' + chrm + '.bim', 'r')
bimer = bim.readlines()
snpids = [x.strip().split()[1] for x in bimer]
snppos = [x.strip().split()[3] for x in bimer]
print "Loading genotypes..."
sys.stdout.flush()
raws = matrix_reader(genodir + 'ByChr/' + outname + '.chr' + chrm + '.raw',
sep=" ")
#linecounter = subprocess.Popen('wc -l ' + genodir + 'ByChr/hutt.imputed.chr' + chrm + '.raw', shell=True, stdout=subprocess.PIPE)
#linecount = int(linecounter.communicate()[0].strip().split()[0])
#columncounter = subprocess.Popen('awk -F" " \'{print NF;exit}\' ' + genodir + 'ByChr/hutt.imputed.chr' + chrm + '.raw', shell=True, stdout=subprocess.PIPE)
#columncount = int(columncounter.communicate()[0].strip().split()[0])
#raws = numpy.zeros((linecount,columncount),dtype='|S2')
#rawin = open(genodir + 'ByChr/hutt.imputed.chr' + chrm + '.raw','r')
#for i,line in enumerate(rawin):
# raws[i,:] = line.strip().split()
#raws = numpy.loadtxt(genodir + 'ByChr/hutt.imputed.chr' + chrm + '.raw',dtype='str')
print "Transposing genotypes..."
sys.stdout.flush()
genorfile = open(genor, 'w')
for line in range(6, raws.shape[1]):