def fitNullTraitByTrait(self,
verbose=True,
cache=False,
out_dir='./cache',
fname=None,
rewrite=False):
"""
Fit null model trait by trait
"""
read_from_file = False
if cache:
assert fname is not None, 'MultiTraitSetTest:: specify fname'
if not os.path.exists(out_dir): os.makedirs(out_dir)
out_file = os.path.join(out_dir, fname)
read_from_file = os.path.exists(out_file) and not rewrite
RV = {}
if read_from_file:
f = h5py.File(out_file, 'r')
for p in range(self.P):
trait_id = self.traitID[p]
g = f[trait_id]
RV[trait_id] = {}
for key in g.keys():
RV[trait_id][key] = g[key][:]
f.close()
self.nullST = RV
else:
""" create mtssST and fit null column by column returns all info """
if self.mtssST is None:
y = SP.zeros((self.N, 1))
self.mtssST = MultiTraitSetTest(y,
XX=self.XX,
S_XX=self.S_XX,
U_XX=self.U_XX,
F=self.F)
RV = {}
for p in range(self.P):
trait_id = self.traitID[p]
y = self.Y[:, p:p + 1]
self.mtssST._setY(y)
RV[trait_id] = self.mtssST.fitNull()
self.nullST = RV
if cache:
f = h5py.File(out_file, 'w')
smartDumpDictHdf5(RV, f)
f.close()
return RV
def fitNullTraitByTrait(self,verbose=True,cache=False,out_dir='./cache',fname=None,rewrite=False):
"""
Fit null model trait by trait
"""
read_from_file = False
if cache:
assert fname is not None, 'MultiTraitSetTest:: specify fname'
if not os.path.exists(out_dir): os.makedirs(out_dir)
out_file = os.path.join(out_dir,fname)
read_from_file = os.path.exists(out_file) and not rewrite
RV = {}
if read_from_file:
f = h5py.File(out_file,'r')
for p in range(self.P):
trait_id = self.traitID[p]
g = f[trait_id]
RV[trait_id] = {}
for key in g.keys():
RV[trait_id][key] = g[key][:]
f.close()
self.nullST=RV
else:
""" create mtssST and fit null column by column returns all info """
if self.mtssST is None:
y = SP.zeros((self.N,1))
self.mtssST = MultiTraitSetTest(y,XX=self.XX,S_XX=self.S_XX,U_XX=self.U_XX,F=self.F)
RV = {}
for p in range(self.P):
trait_id = self.traitID[p]
y = self.Y[:,p:p+1]
self.mtssST._setY(y)
RV[trait_id] = self.mtssST.fitNull()
self.nullST = RV
if cache:
f = h5py.File(out_file,'w')
smartDumpDictHdf5(RV,f)
f.close()
return RV
ipdb.set_trace()
print '.. calculate p-values'
print '(for accurate estimate of pvalues either the number of windows or the number of permutations should be increased)'
c2m = C2M.Chi2mixture(tol=4e-3)
# obtain p-values for mtSet
c2m.estimate_chi2mixture(RV['permutation_llr_mtSet'])
RV['pv_mtSet'] = c2m.sf(RV['llr_mtSet'])
RV['permutation_pv_mtSet'] = c2m.sf(RV['permutation_llr_mtSet'])
# obtain p-values for stSet
RV['pv_stSet'] = SP.zeros_like(RV['llr_stSet'])
RV['permutation_pv_stSet'] = SP.zeros_like(RV['permutation_llr_stSet'])
for p in range(phenotype.shape[1]):
c2m.estimate_chi2mixture(RV['permutation_llr_stSet'][:,p])
RV['pv_stSet'][:,p] = c2m.sf(RV['llr_stSet'][:,p])
RV['permutation_pv_stSet'][:,p] = c2m.sf(RV['permutation_llr_stSet'][:,p])
# obtain p-values for mtSetPC
c2m.estimate_chi2mixture(RV['permutation_llr_mtSetPC'])
RV['pv_mtSetPC'] = c2m.sf(RV['llr_mtSetPC'])
RV['permutation_pv_mtSetPC'] = c2m.sf(RV['permutation_llr_mtSetPC'])
print '.. export results in %s'%files['out_file']
fout = h5py.File(files['out_file'],'w')
smartDumpDictHdf5(RV,fout)
fout.close()
