def test_100genes_main():
opts = {'input': os.path.join(file_dir, 'data/100genes.fa'),
'bed': os.path.join(file_dir, 'data/100genes.bed'),
'mutations': os.path.join(file_dir, 'data/100genes_mutations.txt'),
'output': os.path.join(file_dir, 'output/100genes_deleterious_single_nuc_output.txt'),
'context': 1,
'use_unmapped': False,
'deleterious': 5,
'processes': 0,
'num_iterations': 1000,
'stop_criteria': 100,
'deleterious_pseudo_count': 0,
'unique': False,
'seed': None,
'kind': 'tsg'}
# single nucleotide context
result = pt.main(opts)
num_del_sig = np.sum(result['inactivating BH q-value'] < .1)
assert num_del_sig < 7, 'Few of the 100 test genes should not be significant ({0})'.format(num_del_sig)
# no context case
opts['context'] = 0
opts['output'] = os.path.join(file_dir, 'output/100genes_deleterious_no_context_output.txt')
result = pt.main(opts)
num_del_sig = np.sum(result['inactivating BH q-value'] < .1)
assert num_del_sig < 9, 'Few of the 100 test genes should not be significant ({0})'.format(num_del_sig)
# di-nucleotide context
opts['context'] = 2
opts['output'] = os.path.join(file_dir, 'output/100genes_deleterious_dinuc_output.txt')
result = pt.main(opts)
num_del_sig = np.sum(result['inactivating BH q-value'] < .1)
assert num_del_sig < 7, 'Few of the 100 test genes should not be significant ({0})'.format(num_del_sig)
def test_ctnnb1_hotmaps_main():
opts = {'input': os.path.join(file_dir, 'data/CTNNB1.fa'),
'bed': os.path.join(file_dir, 'data/CTNNB1.bed'),
'mutations': os.path.join(file_dir, 'data/CTNNB1_mutations.txt'),
'output': os.path.join(file_dir, 'output/CTNNB1_output_hotmaps.txt'),
'context': 1.5,
'use_unmapped': False,
'processes': 0,
'num_iterations': 1000,
'stop_criteria': 100,
'unique': 0,
'seed': None,
'window': '3',
'report_index': True,
'null_distr_dir': os.path.join(file_dir, 'output/hotmaps1d_null'),
'kind': 'hotmaps1d'}
# single nucleotide context
result = rt.main(opts)
# di-nucleotide case
opts['window'] = '6'
result = rt.main(opts)
# no context case
opts['window'] = '9'
result = rt.main(opts)
def test_ctnnb1_main():
opts = {'input': os.path.join(file_dir, 'data/CTNNB1.fa'),
'bed': os.path.join(file_dir, 'data/CTNNB1.bed'),
'mutations': os.path.join(file_dir, 'data/CTNNB1_mutations.txt'),
'output': os.path.join(file_dir, 'output/CTNNB1_output.txt'),
'context': 1,
'use_unmapped': False,
'tsg_score': .1,
'recurrent': 3,
'fraction': .02,
'score_dir': os.path.join(file_dir, 'data/scores'),
'processes': 0,
'num_iterations': 10000,
'stop_criteria': 100,
'recurrent_pseudo_count': 0,
'unique': 0,
'seed': None,
'kind': 'oncogene'}
# single nucleotide context
result = pt.main(opts)
assert result.ix[0, 'entropy p-value'] < 0.001, 'CTNNB1 should have a very low p-value ({0}>.001)'.format(result[0][2])
# di-nucleotide case
opts['context'] = 2
result = pt.main(opts)
assert result.ix[0, 'entropy p-value'] < 0.001, 'CTNNB1 should have a very low p-value ({0}>.001)'.format(result[0][2])
# no context case
opts['context'] = 0
result = pt.main(opts)
assert result.ix[0, 'entropy p-value'] < 0.001, 'CTNNB1 should have a very low p-value ({0}>.001)'.format(result[0][2])
def test_tp53_main():
opts = {'input': os.path.join(file_dir, 'data/tp53.fa'),
'bed': os.path.join(file_dir, 'data/tp53.bed'),
'mutations': os.path.join(file_dir, 'data/tp53_mutations.txt'),
'output': os.path.join(file_dir, 'output/tp53_output.txt'),
'context': 1,
'use_unmapped': False,
'deleterious': 5,
'processes': 0,
'num_iterations': 10000,
'stop_criteria': 100,
'deleterious_pseudo_count': 0,
'unique': False,
'seed': None,
'kind': 'tsg'}
# single nucleotide context
result = pt.main(opts)
assert result.ix[0, 'inactivating p-value'] < 0.001, 'TP53 should have a very low p-value ({0}>.001)'.format(result[0][2])
# di-nucleotide case
opts['context'] = 2
result = pt.main(opts)
assert result.ix[0, 'inactivating p-value'] < 0.001, 'TP53 should have a very low p-value ({0}>.001)'.format(result[0][2])
# no context case
opts['context'] = 0
result = pt.main(opts)
assert result.ix[0, 'inactivating p-value'] < 0.001, 'TP53 should have a very low p-value ({0}>.001)'.format(result[0][2])
def test_100genes_main():
opts = {
'input':
os.path.join(file_dir, 'data/100genes.fa'),
'bed':
os.path.join(file_dir, 'data/100genes.bed'),
'mutations':
os.path.join(file_dir, 'data/100genes_mutations.txt'),
'output':
os.path.join(file_dir,
'output/100genes_deleterious_single_nuc_output.txt'),
'context':
1,
'use_unmapped':
False,
'deleterious':
5,
'processes':
0,
'num_iterations':
1000,
'stop_criteria':
100,
'deleterious_pseudo_count':
0,
'unique':
False,
'seed':
None,
'kind':
'tsg'
}
# single nucleotide context
result = pt.main(opts)
num_del_sig = np.sum(result['inactivating BH q-value'] < .1)
assert num_del_sig < 7, 'Few of the 100 test genes should not be significant ({0})'.format(
num_del_sig)
# no context case
opts['context'] = 0
opts['output'] = os.path.join(
file_dir, 'output/100genes_deleterious_no_context_output.txt')
result = pt.main(opts)
num_del_sig = np.sum(result['inactivating BH q-value'] < .1)
assert num_del_sig < 9, 'Few of the 100 test genes should not be significant ({0})'.format(
num_del_sig)
# di-nucleotide context
opts['context'] = 2
opts['output'] = os.path.join(
file_dir, 'output/100genes_deleterious_dinuc_output.txt')
result = pt.main(opts)
num_del_sig = np.sum(result['inactivating BH q-value'] < .1)
assert num_del_sig < 7, 'Few of the 100 test genes should not be significant ({0})'.format(
num_del_sig)
def test_100genes_main():
opts = {
'input':
os.path.join(file_dir, 'data/100genes.fa'),
'bed':
os.path.join(file_dir, 'data/100genes.bed'),
'mutations':
os.path.join(file_dir, 'data/100genes_mutations.txt'),
'output':
os.path.join(file_dir,
'output/100genes_hotmaps_single_nuc_output.txt'),
'context':
1,
'use_unmapped':
False,
'processes':
0,
'num_iterations':
1000,
'stop_criteria':
100,
'unique':
False,
'seed':
None,
'window':
3,
'kind':
'hotmaps1d'
}
# single nucleotide context
result = rt.main(opts)
num_sig = np.sum(result['q-value'] < .01)
assert num_sig < 9, 'Few of the 100 test genes should not be significant ({0})'.format(
num_sig)
def main(opts, mutation_df=None, frameshift_df=None):
# get output file
myoutput_path = opts['output']
opts['output'] = ''
# perform randomization-based test
result_df = rt.main(opts, mutation_df)
# clean up p-values for combined p-value calculation
if opts['kind'] == 'tsg':
p_val_col = 'inactivating p-value'
q_val_col = 'inactivating BH q-value'
elif opts['kind'] == 'effect':
p_val_col = 'entropy-on-effect p-value'
q_val_col = 'entropy-on-effect BH q-value'
elif opts['kind'] == 'oncogene':
p_val_col = 'entropy p-value'
q_val_col = 'entropy BH q-value'
elif opts['kind'] == 'protein':
p_val_col = 'normalized graph-smoothed position entropy p-value'
q_val_col = 'normalized graph-smoothed position entropy BH q-value'
result_df[p_val_col] = result_df[p_val_col].fillna(1)
result_df[q_val_col] = result_df[q_val_col].fillna(1)
if opts['kind'] == 'tsg':
# drop genes that never occur
if opts['kind'] == 'tsg' or opts['kind'] == 'effect':
no_ssvs = (result_df['Total SNV Mutations'] == 0)
result_df = result_df[~no_ssvs]
result_df = result_df.sort_values(by=p_val_col)
elif opts['kind'] == 'oncogene':
# get FDR
result_df = result_df[result_df['Total Mutations'] > 0]
result_df['entropy BH q-value'] = mypval.bh_fdr(
result_df['entropy p-value'])
# combine p-values
result_df['tmp entropy p-value'] = result_df['entropy p-value']
result_df['tmp vest p-value'] = result_df['vest p-value']
result_df.loc[result_df['entropy p-value'] == 0,
'tmp entropy p-value'] = 1. / opts['num_iterations']
result_df.loc[result_df['vest p-value'] == 0,
'tmp vest p-value'] = 1. / opts['num_iterations']
result_df['combined p-value'] = result_df[[
'tmp entropy p-value', 'tmp vest p-value'
]].apply(mypval.fishers_method, axis=1)
result_df['combined BH q-value'] = mypval.bh_fdr(
result_df['combined p-value'])
del result_df['tmp vest p-value']
del result_df['tmp entropy p-value']
if myoutput_path:
# write output if specified
result_df.to_csv(myoutput_path, sep='\t', index=False)
result_df = result_df.set_index('gene', drop=False)
return result_df
def main(opts,
mutation_df=None,
frameshift_df=None):
# get output file
myoutput_path = opts['output']
opts['output'] = ''
# perform randomization-based test
result_df = rt.main(opts, mutation_df)
# clean up p-values for combined p-value calculation
if opts['kind'] == 'tsg':
p_val_col = 'inactivating p-value'
q_val_col = 'inactivating BH q-value'
elif opts['kind'] == 'effect':
p_val_col = 'entropy-on-effect p-value'
q_val_col = 'entropy-on-effect BH q-value'
elif opts['kind'] == 'oncogene':
p_val_col = 'entropy p-value'
q_val_col = 'entropy BH q-value'
elif opts['kind'] == 'protein':
p_val_col = 'normalized graph-smoothed position entropy p-value'
q_val_col = 'normalized graph-smoothed position entropy BH q-value'
elif opts['kind'] == 'hotmaps1d':
p_val_col = 'p-value'
q_val_col = 'q-value'
result_df[p_val_col] = result_df[p_val_col].fillna(1)
result_df[q_val_col] = result_df[q_val_col].fillna(1)
if opts['kind'] == 'tsg':
# drop genes that never occur
if opts['kind'] == 'tsg' or opts['kind'] == 'effect':
no_ssvs = (result_df['Total SNV Mutations']==0)
result_df = result_df[~no_ssvs]
result_df = result_df.sort_values(by=p_val_col)
elif opts['kind'] == 'oncogene':
# get FDR
result_df = result_df[result_df['Total Mutations']>0]
result_df['entropy BH q-value'] = mypval.bh_fdr(result_df['entropy p-value'])
# combine p-values
result_df['tmp entropy p-value'] = result_df['entropy p-value']
result_df['tmp vest p-value'] = result_df['vest p-value']
result_df.loc[result_df['entropy p-value']==0, 'tmp entropy p-value'] = 1. / opts['num_iterations']
result_df.loc[result_df['vest p-value']==0, 'tmp vest p-value'] = 1. / opts['num_iterations']
result_df['combined p-value'] = result_df[['tmp entropy p-value', 'tmp vest p-value']].apply(mypval.fishers_method, axis=1)
result_df['combined BH q-value'] = mypval.bh_fdr(result_df['combined p-value'])
del result_df['tmp vest p-value']
del result_df['tmp entropy p-value']
if myoutput_path:
# write output if specified
result_df.to_csv(myoutput_path, sep='\t', index=False)
result_df = result_df.set_index('gene', drop=False)
return result_df
def test_ctnnb1_main():
opts = {
'input': os.path.join(file_dir, 'data/CTNNB1.fa'),
'bed': os.path.join(file_dir, 'data/CTNNB1.bed'),
'mutations': os.path.join(file_dir, 'data/CTNNB1_mutations.txt'),
'output': os.path.join(file_dir, 'output/CTNNB1_output.txt'),
'context': 1,
'use_unmapped': False,
'tsg_score': .1,
'recurrent': 3,
'fraction': .02,
'score_dir': os.path.join(file_dir, 'data/scores'),
'processes': 0,
'num_iterations': 10000,
'stop_criteria': 100,
'recurrent_pseudo_count': 0,
'unique': 0,
'seed': None,
'kind': 'oncogene'
}
# single nucleotide context
result = pt.main(opts)
assert result.ix[
0,
'entropy p-value'] < 0.001, 'CTNNB1 should have a very low p-value ({0}>.001)'.format(
result[0][2])
# di-nucleotide case
opts['context'] = 2
result = pt.main(opts)
assert result.ix[
0,
'entropy p-value'] < 0.001, 'CTNNB1 should have a very low p-value ({0}>.001)'.format(
result[0][2])
# no context case
opts['context'] = 0
result = pt.main(opts)
assert result.ix[
0,
'entropy p-value'] < 0.001, 'CTNNB1 should have a very low p-value ({0}>.001)'.format(
result[0][2])
def test_tp53_main():
opts = {
'input': os.path.join(file_dir, 'data/tp53.fa'),
'bed': os.path.join(file_dir, 'data/tp53.bed'),
'mutations': os.path.join(file_dir, 'data/tp53_mutations.txt'),
'output': os.path.join(file_dir, 'output/tp53_output.txt'),
'context': 1,
'use_unmapped': False,
'deleterious': 5,
'processes': 0,
'num_iterations': 10000,
'stop_criteria': 100,
'deleterious_pseudo_count': 0,
'unique': False,
'seed': None,
'kind': 'tsg'
}
# single nucleotide context
result = pt.main(opts)
assert result.ix[
0,
'inactivating p-value'] < 0.001, 'TP53 should have a very low p-value ({0}>.001)'.format(
result[0][2])
# di-nucleotide case
opts['context'] = 2
result = pt.main(opts)
assert result.ix[
0,
'inactivating p-value'] < 0.001, 'TP53 should have a very low p-value ({0}>.001)'.format(
result[0][2])
# no context case
opts['context'] = 0
result = pt.main(opts)
assert result.ix[
0,
'inactivating p-value'] < 0.001, 'TP53 should have a very low p-value ({0}>.001)'.format(
result[0][2])
def test_100genes_main():
opts = {'input': os.path.join(file_dir, 'data/100genes.fa'),
'bed': os.path.join(file_dir, 'data/100genes.bed'),
'mutations': os.path.join(file_dir, 'data/100genes_mutations.txt'),
'output': os.path.join(file_dir, 'output/100genes_position_single_nuc_output.txt'),
'context': 1,
'tsg_score': .1,
'recurrent': 3,
'fraction': .02,
'use_unmapped': False,
'processes': 0,
'num_iterations': 1000,
'stop_criteria': 100,
'score_dir': os.path.join(file_dir, 'data/scores'),
'recurrent_pseudo_count': 0,
'unique': False,
'seed': None,
'kind': 'oncogene'}
# single nucleotide context
result = pt.main(opts)
#tested_result = result[result['Performed Recurrency Test']==1]
num_ent_sig = np.sum(result['entropy BH q-value'] < .1)
assert num_ent_sig < 9, 'Few of the 100 test genes should not be significant ({0})'.format(num_ent_sig)
# no context case
opts['context'] = 0
opts['output'] = os.path.join(file_dir, 'output/100genes_position_no_context_output.txt')
result = pt.main(opts)
#tested_result = result[result['Performed Recurrency Test']==1]
num_ent_sig = np.sum(result['entropy BH q-value'] < .1)
assert num_ent_sig < 9, 'Few of the 100 test genes should not be significant ({0})'.format(num_ent_sig)
# di-nucleotide context
opts['context'] = 2
opts['output'] = os.path.join(file_dir, 'output/100genes_position_dinuc_output.txt')
result = pt.main(opts)
#tested_result = result[result['Performed Recurrency Test']==1]
num_ent_sig = np.sum(result['entropy BH q-value'] < .1)
assert num_ent_sig < 9, 'Few of the 100 test genes should not be significant ({0})'.format(num_ent_sig)
def test_100genes_main():
opts = {
'input':
os.path.join(file_dir, 'data/100genes.fa'),
'bed':
os.path.join(file_dir, 'data/100genes.bed'),
'mutations':
os.path.join(file_dir, 'data/100genes_mutations.txt'),
'output':
os.path.join(file_dir,
'output/100genes_position_single_nuc_output.txt'),
'context':
1,
'tsg_score':
.1,
'recurrent':
3,
'fraction':
.02,
'use_unmapped':
False,
'processes':
0,
'num_iterations':
1000,
'stop_criteria':
100,
'score_dir':
os.path.join(file_dir, 'data/scores'),
'recurrent_pseudo_count':
0,
'unique':
False,
'seed':
None,
'kind':
'oncogene'
}
# single nucleotide context
result = pt.main(opts)
#tested_result = result[result['Performed Recurrency Test']==1]
num_ent_sig = np.sum(result['entropy BH q-value'] < .1)
assert num_ent_sig < 9, 'Few of the 100 test genes should not be significant ({0})'.format(
num_ent_sig)
# no context case
opts['context'] = 0
opts['output'] = os.path.join(
file_dir, 'output/100genes_position_no_context_output.txt')
result = pt.main(opts)
#tested_result = result[result['Performed Recurrency Test']==1]
num_ent_sig = np.sum(result['entropy BH q-value'] < .1)
assert num_ent_sig < 9, 'Few of the 100 test genes should not be significant ({0})'.format(
num_ent_sig)
# di-nucleotide context
opts['context'] = 2
opts['output'] = os.path.join(file_dir,
'output/100genes_position_dinuc_output.txt')
result = pt.main(opts)
#tested_result = result[result['Performed Recurrency Test']==1]
num_ent_sig = np.sum(result['entropy BH q-value'] < .1)
assert num_ent_sig < 9, 'Few of the 100 test genes should not be significant ({0})'.format(
num_ent_sig)