import gzip
def custom_db(db_gz, gt_23andMe_gz, n_snps=100000, out_file_gz='out_custom_db.txt.gz'):
'''
+-----------+
| custom_db |
+-----------+
Script used to pull out SNPs present in database and 23andMe file, sort by F_ST, and write top n_snps to gzipped out_file
Package dependencies:
pandas
commanderline
'''
print('Reading database...')
db=pd.read_csv(db_gz, compression='gzip', sep=' ')
print('Reading 23andMe file...')
gt=pd.read_csv(gt_23andMe_gz, compression='gzip', usecols=[1,2], sep='\t', comment='#', header=None)
gt.columns=['chr','pos']
print('Merging on chr&pos...')
m=pd.merge(db, gt, on=['chr','pos'])
print('Sorting by F_ST...')
m=m.sort_values(by='fst', ascending=False)
print('Filtering...')
m_out=m.iloc[:n_snps, :(len(m.columns)-1)]
print('Sorting by chr & pos, and writing to file...')
with gzip.open(out_file_gz, 'wt') as f_out:
m_out.sort_values(by=['chr','pos']).to_csv(f_out, sep=' ', index=False)
cl.commander_line(custom_db) if __name__ == '__main__' else None
if (red == 'avg'):
avg = True
raise Exception(
'Average used, but its output is not always consistent - especially for diag!'
)
elif (red == 'sum'):
avg = False
else:
raise Exception('Error: Unknown argument: ' + red)
# Output is done step-by-step
# analysis.write_output_to_file(out_fname+'.txt.gz', cnst.const['out_delim'], avg)
if generate_img:
analysis.generate_img(out_fname + cnst.const['img_out_ext'])
flat.print_log_msg('Done')
def print_opt_arg_error():
print('For help use --help')
return (2)
# if __name__ == '__main__':
# main()
if __name__ == '__main__':
cl.commander_line((pipeline_lean, pipeline, multiple_pipelines))
commanderline
'''
results=pd.DataFrame()
with open(file_list, 'rt') as f_in:
for i,f in enumerate(f_in):
f=f.strip()
d1=pd.read_csv(f, header=None, sep=' ')
d1['super']=d1[0].apply(lambda el: ancestry_translation[el])
d2=d1.groupby('super').sum()
d2.columns=[i]
results=pd.concat([results, d2], axis=1)
results=results.dropna(subset=[0])
bootstrap_replicate_count=results.iloc[:,1:].apply(lambda el: el>=negligible_threshold).T.sum()/(len(results.columns)-1)
filt=bootstrap_replicate_count.apply(lambda el: el>=bootstrap_confidence)
interm_results=results[filt][0]
ambig=1-interm_results.sum()
final_results=pd.concat([interm_results, pd.Series({'AMBIGUOUS':ambig})])
print(final_results)
cl.commander_line(validate_bootstrap) if __name__ == '__main__' else None
def calc_fst(ref_gz, out_gz='out_db_fst.txt.gz'):
'''
+----------+
| calc_fst |
+----------+
Helper script to calculate F_ST for Ancestry-format reference files. Output is the input file with an additional column named 'fst' (SNPs with NaN for F_ST are removed)
Details about Ancestry:
https://bitbucket.org/joepickrell/ancestry
Package dependencies:
pandas
commanderline
'''
print('Reading Ancestry-format reference file...')
d = pd.read_csv(ref_gz, sep=' ')
# Drop null columns, e.g., last column followed by a column delimiter will result in a null column
d = d.dropna(axis=1, how='all')
print('Calculating F_ST...')
d['fst'] = d.iloc[:, 6:].apply(fst, axis=1)
# Drop SNPs without F_ST
d = d.dropna(subset=['fst'])
print('Writing output file...')
with gzip.open(out_gz, 'wt') as f_out:
d.to_csv(f_out, sep=' ', index=False)
cl.commander_line(calc_fst) if __name__ == '__main__' else None
with open(file_list, 'rt') as f_in:
for i, f in enumerate(f_in):
f = f.strip()
d1 = pd.read_csv(f, header=None, sep=' ')
d1['super'] = d1[0].apply(lambda el: ancestry_translation[el])
d2 = d1.groupby('super').sum()
d2.columns = [i]
results = pd.concat([results, d2], axis=1)
results = results.dropna(subset=[0])
bootstrap_replicate_count = results.iloc[:, 1:].apply(
lambda el: el >= negligible_threshold).T.sum() / (
len(results.columns) - 1)
filt = bootstrap_replicate_count.apply(
lambda el: el >= bootstrap_confidence)
interm_results = results[filt][0]
ambig = 1 - interm_results.sum()
final_results = pd.concat(
[interm_results, pd.Series({'AMBIGUOUS': ambig})])
print(final_results)
cl.commander_line(validate_bootstrap) if __name__ == '__main__' else None
def calc_fst(ref_gz, out_gz='out_db_fst.txt.gz'):
'''
+----------+
| calc_fst |
+----------+
Helper script to calculate F_ST for Ancestry-format reference files. Output is the input file with an additional column named 'fst' (SNPs with NaN for F_ST are removed)
Details about Ancestry:
https://bitbucket.org/joepickrell/ancestry
Package dependencies:
pandas
commanderline
'''
print('Reading Ancestry-format reference file...')
d=pd.read_csv(ref_gz, sep=' ')
# Drop null columns, e.g., last column followed by a column delimiter will result in a null column
d=d.dropna(axis=1,how='all')
print('Calculating F_ST...')
d['fst']=d.iloc[:,6:].apply(fst, axis=1)
# Drop SNPs without F_ST
d=d.dropna(subset=['fst'])
print('Writing output file...')
with gzip.open(out_gz, 'wt') as f_out:
d.to_csv(f_out, sep=' ', index=False)
cl.commander_line(calc_fst) if __name__ == '__main__' else None
Script used to pull out SNPs present in database and 23andMe file, sort by F_ST, and write top n_snps to gzipped out_file
Package dependencies:
pandas
commanderline
'''
print('Reading database...')
db = pd.read_csv(db_gz, compression='gzip', sep=' ')
print('Reading 23andMe file...')
gt = pd.read_csv(gt_23andMe_gz,
compression='gzip',
usecols=[1, 2],
sep='\t',
comment='#',
header=None)
gt.columns = ['chr', 'pos']
print('Merging on chr&pos...')
m = pd.merge(db, gt, on=['chr', 'pos'])
print('Sorting by F_ST...')
m = m.sort_values(by='fst', ascending=False)
print('Filtering...')
m_out = m.iloc[:n_snps, :(len(m.columns) - 1)]
print('Sorting by chr & pos, and writing to file...')
with gzip.open(out_file_gz, 'wt') as f_out:
m_out.sort_values(by=['chr', 'pos']).to_csv(f_out,
sep=' ',
index=False)
cl.commander_line(custom_db) if __name__ == '__main__' else None
print_metric(metric1)
flat.print_log_msg('Breakpoint 2: ' + repr(breakpoint2))
flat.print_log_msg('Metric 2:')
print_metric(metric2)
def apply_metric(chr_name, begin, end, input_config, loci):
metric_out = metric.Metric(chr_name, input_config, loci, begin, end)
out = metric_out.calc_metric()
return out
def print_metric(metric_out):
flat.print_log_msg('Sum: ' + repr(metric_out['sum']))
flat.print_log_msg('N (w/ zero\'s): ' + repr(metric_out['N_zero']))
flat.print_log_msg('Metric: ' +
repr(metric_out['sum'] / metric_out['N_zero']))
def print_opt_arg_error():
print('For help use --help')
return (2)
# if __name__ == '__main__':
# main()
if __name__ == '__main__':
cl.commander_line((pipeline))
dist = -0.1
for key in loaded:
# print()
# print(key, loaded[key])
try:
if 'loci' in loaded[key]:
dist -= 0.05
breakpoint_list_y = [dist for a in loaded[key]['loci']]
pt.scatter(np.array(loaded[key]['loci']),
np.array(breakpoint_list_y),
c=next(colors),
label=key,
marker=next(markers))
except TypeError as e:
print(e)
pt.legend()
pt.show()
def plot_gwas_w_default_paths(chr_name, input_pickle_fname):
plot_gwas(
cnst.const['gwas_plots']['height']['root'] +
cnst.const['gwas_plots']['height']['file_prefix'] + chr_name +
cnst.const['gwas_plots']['height']['file_suffix'], input_pickle_fname)
if __name__ == '__main__':
cl.commander_line((plot_gwas_w_default_paths, plot_gwas))
# input_config = cnst.const['orig_data_'+dataset]
input_config = cnst.return_conf(dataset_path)
partitions = flat.read_partitions(name, input_config)
with open(input_pickle_fname, 'rb') as f_in:
loaded = pickle.load(f_in)
# print(loaded)
loci = loaded[subset]['loci']
first = partitions[0][0]
last = partitions[len(partitions) - 1][1]
# print(loci)
print('chr', '\t', 'start', '\t', 'stop')
print(name, '\t', first, '\t', loci[0])
for i in range(0, len(loci) - 1):
print(name, '\t', loci[i], '\t', loci[i + 1])
print(name, '\t', loci[len(loci) - 1], '\t', last + 1)
if __name__ == '__main__':
cl.commander_line((chr_bpoints_to_bed, ),
print_argv_to_output=False,
print_done=False)
if (red == 'avg'):
avg = True
raise Exception(
'Average used, but its output is not always consistent - especially for diag!'
)
elif (red == 'sum'):
avg = False
else:
raise Exception('Error: Unknown argument: ' + red)
# Output is done step-by-step
# analysis.write_output_to_file(out_fname+'.txt.gz', cnst.const['out_delim'], avg)
if generate_img:
analysis.generate_img(out_fname + cnst.const['img_out_ext'])
flat.print_log_msg('Done')
def print_opt_arg_error():
print('For help use --help')
return (2)
# if __name__ == '__main__':
# main()
if __name__ == '__main__':
cl.commander_line((pipeline_lean, ))
