def computePCsPython(out_dir,k,bfile,ffile):
""" reading in """
RV = plink_reader.readBED(bfile,useMAFencoding=True)
X = RV['snps']
""" normalizing markers """
print 'Normalizing SNPs...'
p_ref = X.mean(axis=0)/2.
X -= 2*p_ref
with warnings.catch_warnings():
warnings.simplefilter("ignore")
X /= SP.sqrt(2*p_ref*(1-p_ref))
hasNan = SP.any(SP.isnan(X),axis=0)
print '%d SNPs have a nan entry. Exluding them for computing the covariance matrix.'%hasNan.sum()
X = X[:,~hasNan]
""" computing prinicipal components """
U,S,Vt = SSL.svds(X,k=k)
U -= U.mean(0)
U /= U.std(0)
U = U[:,::-1]
""" saving to output """
NP.savetxt(ffile, U, delimiter='\t',fmt='%.6f')
def getRegion(self,
size=3e4,
min_nSNPs=1,
chrom_i=None,
pos_min=None,
pos_max=None):
"""
Sample a region from the piece of genotype X, chrom, pos
minSNPnum: minimum number of SNPs contained in the region
Ichrom: restrict X to chromosome Ichrom before taking the region
cis: bool vector that marks the sorted region
region: vector that contains chrom and init and final position of the region
"""
bim = plink_reader.readBIM(self.bfile, usecols=(0, 1, 2, 3))
chrom = SP.array(bim[:, 0], dtype=int)
pos = SP.array(bim[:, 3], dtype=int)
if chrom_i is None:
n_chroms = chrom.max()
chrom_i = int(SP.ceil(SP.rand() * n_chroms))
pos = pos[chrom == chrom_i]
chrom = chrom[chrom == chrom_i]
ipos = SP.ones(len(pos), dtype=bool)
if pos_min is not None:
ipos = SP.logical_and(ipos, pos_min < pos)
if pos_max is not None:
ipos = SP.logical_and(ipos, pos < pos_max)
pos = pos[ipos]
chrom = chrom[ipos]
if size == 1:
# select single SNP
idx = int(SP.ceil(pos.shape[0] * SP.rand()))
cis = SP.arange(pos.shape[0]) == idx
region = SP.array([chrom_i, pos[idx], pos[idx]])
else:
while 1:
idx = int(SP.floor(pos.shape[0] * SP.rand()))
posT1 = pos[idx]
posT2 = pos[idx] + size
if posT2 <= pos.max():
cis = chrom == chrom_i
cis *= (pos > posT1) * (pos < posT2)
if cis.sum() > min_nSNPs: break
region = SP.array([chrom_i, posT1, posT2])
start = SP.nonzero(cis)[0].min()
nSNPs = cis.sum()
rv = plink_reader.readBED(self.bfile,
useMAFencoding=True,
start=start,
nSNPs=nSNPs,
bim=bim)
Xr = rv['snps']
return Xr, region
def scan(bfile,Y,cov,null,wnds,minSnps,i0,i1,perm_i,resfile,F):
if perm_i is not None:
print 'Generating permutation (permutation %d)'%perm_i
NP.random.seed(perm_i)
perm = NP.random.permutation(Y.shape[0])
mtSet = MTST.MultiTraitSetTest(Y,S_XX=cov['eval'],U_XX=cov['evec'],F=F)
mtSet.setNull(null)
bim = plink_reader.readBIM(bfile,usecols=(0,1,2,3))
fam = plink_reader.readFAM(bfile,usecols=(0,1))
print 'fitting model'
wnd_file = csv.writer(open(resfile,'wb'),delimiter='\t')
for wnd_i in range(i0,i1):
print '.. window %d - (%d, %d-%d) - %d snps'%(wnd_i,int(wnds[wnd_i,1]),int(wnds[wnd_i,2]),int(wnds[wnd_i,3]),int(wnds[wnd_i,-1]))
if int(wnds[wnd_i,-1])<minSnps:
print 'SKIPPED: number of snps lower than minSnps'
continue
#RV = bed.read(PositionRange(int(wnds[wnd_i,-2]),int(wnds[wnd_i,-1])))
RV = plink_reader.readBED(bfile, useMAFencoding=True, blocksize = 1, start = int(wnds[wnd_i,4]), nSNPs = int(wnds[wnd_i,5]), order = 'F',standardizeSNPs=False,ipos = 2,bim=bim,fam=fam)
Xr = RV['snps']
if perm_i is not None:
Xr = Xr[perm,:]
rv = mtSet.optimize(Xr)
line = NP.concatenate([wnds[wnd_i,:],rv['LLR']])
wnd_file.writerow(line)
pass
def getRegion(self, size=3e4, min_nSNPs=1, chrom_i=None, pos_min=None, pos_max=None):
"""
Sample a region from the piece of genotype X, chrom, pos
minSNPnum: minimum number of SNPs contained in the region
Ichrom: restrict X to chromosome Ichrom before taking the region
cis: bool vector that marks the sorted region
region: vector that contains chrom and init and final position of the region
"""
bim = plink_reader.readBIM(self.bfile, usecols=(0, 1, 2, 3))
chrom = SP.array(bim[:, 0], dtype=int)
pos = SP.array(bim[:, 3], dtype=int)
if chrom_i is None:
n_chroms = chrom.max()
chrom_i = int(SP.ceil(SP.rand() * n_chroms))
pos = pos[chrom == chrom_i]
chrom = chrom[chrom == chrom_i]
ipos = SP.ones(len(pos), dtype=bool)
if pos_min is not None:
ipos = SP.logical_and(ipos, pos_min < pos)
if pos_max is not None:
ipos = SP.logical_and(ipos, pos < pos_max)
pos = pos[ipos]
chrom = chrom[ipos]
if size == 1:
# select single SNP
idx = int(SP.ceil(pos.shape[0] * SP.rand()))
cis = SP.arange(pos.shape[0]) == idx
region = SP.array([chrom_i, pos[idx], pos[idx]])
else:
while 1:
idx = int(SP.floor(pos.shape[0] * SP.rand()))
posT1 = pos[idx]
posT2 = pos[idx] + size
if posT2 <= pos.max():
cis = chrom == chrom_i
cis *= (pos > posT1) * (pos < posT2)
if cis.sum() > min_nSNPs:
break
region = SP.array([chrom_i, posT1, posT2])
start = SP.nonzero(cis)[0].min()
nSNPs = cis.sum()
rv = plink_reader.readBED(self.bfile, useMAFencoding=True, start=start, nSNPs=nSNPs, bim=bim)
Xr = rv["snps"]
return Xr, region
def _genBgTerm_fromSNPs(self,
vTot=0.5,
vCommon=0.1,
pCausal=0.5,
plot=False):
""" generate """
print 'Reading in all SNPs. This is slow.'
rv = plink_reader.readBED(self.bfile, useMAFencoding=True)
X = rv['snps']
S = X.shape[1]
vSpecific = vTot - vCommon
# select causal SNPs
nCausal = int(SP.floor(pCausal * S))
Ic = selectRnd(nCausal, S)
X = X[:, Ic]
# common effect
Bc = SP.kron(SP.randn(nCausal, 1), SP.randn(1, self.P))
Yc = SP.dot(X, Bc)
Yc *= SP.sqrt(vCommon / Yc.var(0).mean())
# indipendent effect
Bi = SP.randn(nCausal, self.P)
Yi = SP.dot(X, Bi)
Yi *= SP.sqrt(vSpecific / Yi.var(0).mean())
if plot:
import pylab as PL
PL.ion()
for p in range(self.P):
PL.subplot(self.P, 1, p + 1)
PL.plot(SP.arange(self.X.shape[1])[Ic],
Bc[:, p],
'o',
color='y',
alpha=0.05)
PL.plot(SP.arange(self.X.shape[1])[Ic],
Bi[:, p],
'o',
color='r',
alpha=0.05)
#PL.ylim(-2,2)
PL.plot([0, Ic.shape[0]], [0, 0], 'k')
return Yc, Yi
def scan(bfile, Y, cov, null, wnds, minSnps, i0, i1, perm_i, resfile, F):
if perm_i is not None:
print 'Generating permutation (permutation %d)' % perm_i
NP.random.seed(perm_i)
perm = NP.random.permutation(Y.shape[0])
mtSet = MTST.MultiTraitSetTest(Y, S_XX=cov['eval'], U_XX=cov['evec'], F=F)
mtSet.setNull(null)
bim = plink_reader.readBIM(bfile, usecols=(0, 1, 2, 3))
fam = plink_reader.readFAM(bfile, usecols=(0, 1))
print 'fitting model'
wnd_file = csv.writer(open(resfile, 'wb'), delimiter='\t')
for wnd_i in range(i0, i1):
print '.. window %d - (%d, %d-%d) - %d snps' % (
wnd_i, int(wnds[wnd_i, 1]), int(wnds[wnd_i, 2]), int(
wnds[wnd_i, 3]), int(wnds[wnd_i, -1]))
if int(wnds[wnd_i, -1]) < minSnps:
print 'SKIPPED: number of snps lower than minSnps'
continue
#RV = bed.read(PositionRange(int(wnds[wnd_i,-2]),int(wnds[wnd_i,-1])))
RV = plink_reader.readBED(bfile,
useMAFencoding=True,
blocksize=1,
start=int(wnds[wnd_i, 4]),
nSNPs=int(wnds[wnd_i, 5]),
order='F',
standardizeSNPs=False,
ipos=2,
bim=bim,
fam=fam)
Xr = RV['snps']
if perm_i is not None:
Xr = Xr[perm, :]
rv = mtSet.optimize(Xr)
line = NP.concatenate([wnds[wnd_i, :], rv['LLR']])
wnd_file.writerow(line)
pass
def computeCovarianceMatrixPython(out_dir,bfile,cfile,sim_type='RRM'):
print "Using python to create covariance matrix. This might be slow. We recommend using plink instead."
if sim_type is not 'RRM':
raise Exception('sim_type %s is not known'%sim_type)
""" loading data """
data = plink_reader.readBED(bfile,useMAFencoding=True)
iid = data['iid']
X = data['snps']
N = X.shape[1]
print '%d variants loaded.'%N
print '%d people loaded.'%X.shape[0]
""" normalizing markers """
print 'Normalizing SNPs...'
p_ref = X.mean(axis=0)/2.
X -= 2*p_ref
with warnings.catch_warnings():
warnings.simplefilter("ignore")
X /= SP.sqrt(2*p_ref*(1-p_ref))
hasNan = SP.any(SP.isnan(X),axis=0)
print '%d SNPs have a nan entry. Exluding them for computing the covariance matrix.'%hasNan.sum()
""" computing covariance matrix """
print 'Computing relationship matrix...'
K = SP.dot(X[:,~hasNan],X[:,~hasNan].T)
K/= 1.*N
print 'Relationship matrix calculation complete'
print 'Relationship matrix written to %s.cov.'%cfile
print 'IDs written to %s.cov.id.'%cfile
""" saving to output """
NP.savetxt(cfile + '.cov', K, delimiter='\t',fmt='%.6f')
NP.savetxt(cfile + '.cov.id', iid, delimiter=' ',fmt='%s')
def _genBgTerm_fromSNPs(self, vTot=0.5, vCommon=0.1, pCausal=0.5, plot=False):
""" generate """
print "Reading in all SNPs. This is slow."
rv = plink_reader.readBED(self.bfile, useMAFencoding=True)
X = rv["snps"]
S = X.shape[1]
vSpecific = vTot - vCommon
# select causal SNPs
nCausal = int(SP.floor(pCausal * S))
Ic = selectRnd(nCausal, S)
X = X[:, Ic]
# common effect
Bc = SP.kron(SP.randn(nCausal, 1), SP.randn(1, self.P))
Yc = SP.dot(X, Bc)
Yc *= SP.sqrt(vCommon / Yc.var(0).mean())
# indipendent effect
Bi = SP.randn(nCausal, self.P)
Yi = SP.dot(X, Bi)
Yi *= SP.sqrt(vSpecific / Yi.var(0).mean())
if plot:
import pylab as PL
PL.ion()
for p in range(self.P):
PL.subplot(self.P, 1, p + 1)
PL.plot(SP.arange(self.X.shape[1])[Ic], Bc[:, p], "o", color="y", alpha=0.05)
PL.plot(SP.arange(self.X.shape[1])[Ic], Bi[:, p], "o", color="r", alpha=0.05)
# PL.ylim(-2,2)
PL.plot([0, Ic.shape[0]], [0, 0], "k")
return Yc, Yi
