def __save_and_load_problem_npz(self, problem):
'''Save and load a problem from NPZ file.'''
out_file = tempfile.TemporaryFile()
io.write_npz(problem, out_file)
# Only needed here to simulate closing & reopening file; you will need to call
# out_file.close() on to prevent file locking in Windows
out_file.seek(0)
return io.read_npz(out_file)
def __npz_to_plink(self, p, file_name):
'''Convert p from npz to plink format using the file set specified by file_name.'''
npz = file_name+'.npz'
# Save test problem in plink format
io.write_npz(p, npz)
# Convert plink -> npz
io.npz_to_plink(npz, file_name)
# Load npz and check that the problem object didn't change
p2 = io.read_plink(prefix=file_name)
return p2
def __main(options):
'''
--------------------------------------------------
Main program - accepts an options struct.
--------------------------------------------------
'''
if options.debug: print 'Input options', options
print 'Building phaser (stage = %d) ...' % (options.stage, )
phaser = build_phasing_pipeline(options)
if options.debug: print 'Reading data ...'
problem = __load_problem(options)
if options.debug: print 'Phasing ...'
params = PhaseParam()
params.update_from_struct(options)
request = run_phasing_chain(phaser, problem, params)
print ''
request.stats.pprint()
print ''
if options.output is not None:
if options.min_output:
print 'Minimizing output size...'
io.slim(problem)
out_prefix, ext = os.path.splitext(options.output)
if ext == '.npz':
print 'Writing haplotype result to %s in NPZ format ...' % (
options.output, )
io.write_npz(problem, options.output)
output_info = out_prefix + '.info.npz'
print 'Writing problem info result to %s in NPZ format ...' % (
output_info, )
io.write_info_npz(problem.info, output_info)
else:
print 'Writing haplotype result to %s in PLINK format ...' % (
options.output, )
io.write_plink(problem, options.output, verbose=options.debug)
return problem
def __main(options):
'''
--------------------------------------------------
Main program - accepts an options struct.
--------------------------------------------------
'''
if options.debug: print 'Input options', options
print 'Building phaser (stage = %d) ...' % (options.stage,)
phaser = build_phasing_pipeline(options)
if options.debug: print 'Reading data ...'
problem = __load_problem(options)
if options.debug: print 'Phasing ...'
params = PhaseParam()
params.update_from_struct(options)
request = run_phasing_chain(phaser, problem, params)
print ''
request.stats.pprint()
print ''
if options.output is not None:
if options.min_output:
print 'Minimizing output size...'
io.slim(problem)
out_prefix, ext = os.path.splitext(options.output)
if ext == '.npz':
print 'Writing haplotype result to %s in NPZ format ...' % (options.output,)
io.write_npz(problem, options.output)
output_info = out_prefix + '.info.npz'
print 'Writing problem info result to %s in NPZ format ...' % (output_info,)
io.write_info_npz(problem.info, output_info)
else:
print 'Writing haplotype result to %s in PLINK format ...' % (options.output,)
io.write_plink(problem, options.output, verbose=options.debug)
return problem
util.run_command('%s --nonfounders --freq --out %s' % (plink_cmd_base, out_base_name))
# Convert frequencies file that to a reference allele recoding
# file (a file containing the list of SNPs and their minor allele letter)
bu.frq_to_minor_file(out_base_name + '.frq', out_base_name + '.mnr')
# Then convert binary PLINK to a recoded 12-recoded TPED, where 1=minor allele for each SNP
out_recoded = out_base_name + '.recoded'
util.run_command('%s --transpose --recode12 --reference-allele %s.mnr --out %s' % \
(plink_cmd_base, out_base_name, out_recoded))
# Reload the recoded problem
for ext in ('nof', 'tped', 'tfam'):
os.rename(out_recoded + '.' + ext, out_base_name + '.' + ext)
genotype = io_genotype.read('plink', 'genotype', tped=out_base_name + '.tped', load_ids=False)
else:
genotype = problem.genotype
# Write problem to file in our (npz)
io.write_npz(problem, out_base_name + '.npz')
# Write genotypes Gaixin formats; she uses those separate files
io_genotype.write('gaixin', genotype, options.out_gxn + '.gxn',
sample_id=problem.pedigree.sample_id_genotyped)
# Convert plink tped to bed; delete the tped set
util.run_command('%s --make-bed --out %s' % (plink_cmd_base, out_base_name))
for ext in ('nof', 'pdg.tfam', 'tped', 'tfam', 'info'):
os.remove(out_base_name + '.' + ext)
except:
traceback.print_exc(file=sys.stdout)
sys.exit(util.EXIT_FAILURE)