def get_mutation_profile_ppi(data_folder, ssc_mutation_data, ssc_subgroups,
gene_data, result_folder, mut_type, alpha, tol,
lambd, influence_weight, simplification,
keep_singletons, ngh_max):
if mut_type == 'raw':
mut_propag, mp_gene, mp_indiv = (
load_data.load_specific_SSC_mutation_profile(
data_folder, ssc_mutation_data, ssc_subgroups, gene_data))
else:
final_influence_mutation_directory = result_folder + 'final_influence/'
final_influence_mutation_file = (
final_influence_mutation_directory +
'final_influence_mutation_profile_{}_alpha={}_tol={}.mat'.format(
mut_type, alpha, tol))
final_influence_data = loadmat(final_influence_mutation_file)
mut_propag = final_influence_data['mut_propag']
if lambd != 0:
ppi_final_file = (
final_influence_mutation_directory +
'PPI_final_weight={}_simp={}_alpha={}_tol={}_singletons={}_ngh={}.mat'
.format(influence_weight, simplification, alpha, tol,
keep_singletons, ngh_max))
ppi_final_data = loadmat(ppi_final_file)
ppi_final = ppi_final_data['ppi_final']
return mut_propag, ppi_final
return mut_propag
def preprocessing(ppi_data,
mut_type,
ssc_subgroups,
data_folder,
patient_data,
ssc_mutation_data,
gene_data,
influence_weight,
simplification,
compute,
overwrite,
tol,
ngh_max,
keep_singletons,
min_mutation,
max_mutation,
result_folder,
alpha,
return_val=False):
if mut_type == 'raw':
mutation_profile, mp_gene, mp_indiv = (
load_data.load_specific_SSC_mutation_profile(
data_folder, ssc_mutation_data, ssc_subgroups, gene_data))
else:
gene_id, mp_indiv, entrez_ppi, idx_filtred, mutation_profile = preprocess_noRaw(
ppi_data, mut_type, ssc_subgroups, data_folder, patient_data,
ssc_mutation_data, gene_data, influence_weight, simplification,
compute, overwrite, tol, ngh_max, keep_singletons, min_mutation,
max_mutation, result_folder, alpha)
# Entrez Gene ID in filtered/formatted mutation profile
mp_gene = [entrez_ppi[i] for i in idx_filtred]
if return_val:
return mutation_profile, mp_indiv, mp_gene, entrez_ppi, idx_filtred
def preprocess_noRaw(ppi_data, mut_type, ssc_subgroups, data_folder,
patient_data, ssc_mutation_data, gene_data,
influence_weight, simplification, compute, overwrite, tol,
ngh_max, keep_singletons, min_mutation, max_mutation,
result_folder, alpha):
print("------------ load_data.py ------------ {}".format(
datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S')),
flush=True)
if patient_data == 'TCGA_UCEC':
(individual_id, mutation_profile, gene_id, gene_symbol_profile) = (
load_data.load_TCGA_UCEC_patient_data(data_folder))
elif patient_data == 'Faroe':
mutation_profile, gene_id = (
load_data.load_Faroe_Islands_data(data_folder))
elif patient_data == 'SSC':
mutation_profile, gene_id, individual_id = (
load_data.load_specific_SSC_mutation_profile(
data_folder, ssc_mutation_data, ssc_subgroups, gene_data))
if ppi_data == 'Hofree_STRING':
gene_id_ppi, network = load_data.load_Hofree_PPI_String(
data_folder, ppi_data)
else:
gene_id_ppi, network = load_data.load_PPI_network(
data_folder, ppi_data)
print("------------ formatting_data.py ------------ {}".format(
datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S')),
flush=True)
idx_ppi, idx_ppi_only, ppi_total, mut_total, ppi_filt = (
formatting_data.formatting(network, mutation_profile, gene_id_ppi,
gene_id))
# EntrezGene ID to int
entrez_ppi = [int(i) for i in gene_id_ppi]
# EntrezGene indexes in PPI after formatting
idx_filtred = idx_ppi + idx_ppi_only
print("------------ filtering_diffusion.py ------------ {}".format(
datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S')),
flush=True)
ppi_final, mut_propag = (filtering_diffusion.filtering(
ppi_filt, result_folder, influence_weight, simplification, compute,
overwrite, alpha, tol, ppi_total, mut_total, ngh_max, keep_singletons,
min_mutation, max_mutation, mut_type))
return gene_id, individual_id, entrez_ppi, idx_filtred, mut_propag
def preprocessing(data_folder, patient_data, ssc_mutation_data, ssc_subgroups, gene_data,
ppi_data, result_folder, influence_weight, simplification,
compute, overwrite, alpha, tol, ngh_max,
keep_singletons, min_mutation, max_mutation, mut_type):
print("------------ load_data.py ------------ {}"
.format(datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S')),
flush=True)
if patient_data == 'TCGA_UCEC':
(patient_id, mutation_profile, gene_id_patient, gene_symbol_profile) = (
load_data.load_TCGA_UCEC_patient_data(data_folder))
elif patient_data == 'Faroe':
mutation_profile, gene_id_patient = (
load_data.load_Faroe_Islands_data(data_folder))
elif patient_data == 'SSC':
mutation_profile, gene_id_patient, patient_id = (
load_data.load_specific_SSC_mutation_profile(
data_folder, ssc_mutation_data, ssc_subgroups, gene_data))
if ppi_data == 'Hofree_STRING':
gene_id_ppi, network = load_data.load_Hofree_PPI_String(
data_folder, ppi_data)
else:
gene_id_ppi, network = load_data.load_PPI_network(
data_folder, ppi_data)
print("------------ formatting_data.py ------------ {}"
.format(datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S')),
flush=True)
idx_ppi, idx_ppi_only, ppi_total, mut_total, ppi_filt = (
formatting_data.formatting(
network, mutation_profile, gene_id_ppi, gene_id_patient))
print("------------ filtering_diffusion.py ------------ {}"
.format(datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S')),
flush=True)
ppi_final, mut_propag = (
filtering_diffusion.filtering(
ppi_filt, result_folder, influence_weight, simplification, compute,
overwrite, alpha, tol, ppi_total, mut_total, ngh_max,
keep_singletons, min_mutation, max_mutation, mut_type))
return gene_id_ppi, idx_ppi, idx_ppi_only
def all_functions(params):
if alpha == 0 and qn is not None:
print('############ PASS ############')
pass
else:
if patient_data == 'SSC':
result_folder = (data_folder + 'result_' + ssc_mutation_data + '_' +
ssc_subgroups + '_' + gene_data + '_' + ppi_data + '/')
else:
result_folder = (data_folder + 'result_' + patient_data + '_' +
ppi_data + '/')
print(result_folder, flush=True)
print("alpha =", alpha, flush=True)
print("QN =", qn, flush=True)
print("k =", n_components, flush=True)
print("lambda =", lambd, flush=True)
print("PPI network =", ppi_data, flush=True)
# ------------ load_data.py ------------
print("------------ load_data.py ------------ {}"
.format(datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S')), flush=True)
if patient_data == 'TCGA_UCEC':
(patient_id, mutation_profile, gene_id_patient,
gene_symbol_profile) = load_data.load_TCGA_UCEC_patient_data(
data_folder)
elif patient_data == 'Faroe':
mutation_profile, gene_id_patient = (
load_data.load_Faroe_Islands_data(data_folder))
elif patient_data == 'SSC':
mutation_profile, gene_id_patient, patient_id = (
load_data.load_specific_SSC_mutation_profile(
data_folder, ssc_mutation_data, ssc_subgroups, gene_data))
if ppi_data == 'Hofree_STRING':
gene_id_ppi, network = load_data.load_Hofree_PPI_String(
data_folder, ppi_data)
else:
gene_id_ppi, network = load_data.load_PPI_network(
data_folder, ppi_data)
# ------------ formatting_data.py ------------
print("------------ formatting_data.py ------------ {}"
.format(datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S')), flush=True)
(network, mutation_profile,
idx_ppi, idx_mut, idx_ppi_only, idx_mut_only) = (
formatting_data.classify_gene_index(
network, mutation_profile, gene_id_ppi, gene_id_patient))
(ppi_total, mut_total, ppi_filt, mut_filt) = (
formatting_data.all_genes_in_submatrices(
network, idx_ppi, idx_mut, idx_ppi_only, idx_mut_only,
mutation_profile))
# ------------ filtering_diffusion.py ------------
print("------------ filtering_diffusion.py ------------ {}"
.format(datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S')), flush=True)
final_influence = (
filtering_diffusion.calcul_final_influence(
sp.eye(ppi_filt.shape[0], dtype=np.float32), ppi_filt,
result_folder, influence_weight, simplification,
compute, overwrite, alpha, tol))
ppi_final, mut_final = filtering_diffusion.filter_ppi_patients(
ppi_total, mut_total, ppi_filt, final_influence, ngh_max,
keep_singletons, min_mutation, max_mutation)
mut_type, mut_propag = filtering_diffusion.propagation_profile(
mut_final, ppi_filt, result_folder, alpha, tol, qn)
# ------------ clustering.py ------------
print("------------ clustering.py ------------ {}"
.format(datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S')), flush=True)
genes_clustering, patients_clustering = (clustering.bootstrap(
result_folder, mut_type, mut_propag, ppi_final,
influence_weight, simplification,
alpha, tol, keep_singletons, ngh_max, min_mutation, max_mutation,
n_components, n_permutations,
run_bootstrap, lambd, tol_nmf, compute_gene_clustering))
distance_genes, distance_patients = clustering.consensus_clustering(
result_folder, genes_clustering, patients_clustering,
influence_weight, simplification, mut_type,
alpha, tol, keep_singletons, ngh_max, min_mutation, max_mutation,
n_components, n_permutations, run_consensus, lambd, tol_nmf,
compute_gene_clustering)
# ------------ hierarchical_clustering.py ------------
print("------------ hierarchical_clustering.py ------------ {}"
.format(datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S')), flush=True)
# if alpha > 0:
# if qn == 'mean':
# mut_type = 'mean_qn'
# elif qn == 'median':
# mut_type = 'median_qn'
# else:
# mut_type = 'diff'
# else:
# mut_type = 'raw'
# print("mutation type =", mut_type)
#
# consensus_directory = result_folder+'consensus_clustering/'
# consensus_mut_type_directory = consensus_directory + mut_type + '/'
#
# hierarchical_directory = result_folder+'hierarchical_clustering/'
# os.makedirs(hierarchical_directory, exist_ok=True)
# hierarchical_mut_type_directory = hierarchical_directory + mut_type + '/'
# os.makedirs(hierarchical_mut_type_directory, exist_ok=True)
#
# if lambd > 0:
# consensus_factorization_directory = (consensus_mut_type_directory + 'gnmf/')
# hierarchical_factorization_directory = (hierarchical_mut_type_directory + 'gnmf/')
#
# else:
# consensus_factorization_directory = (consensus_mut_type_directory + 'nmf/')
# hierarchical_factorization_directory = (hierarchical_mut_type_directory + 'nmf/')
# os.makedirs(hierarchical_factorization_directory, exist_ok=True)
#
# consensus_file = (consensus_factorization_directory +
# 'consensus_weight={}_simp={}_alpha={}_tol={}_singletons={}_ngh={}_minMut={}_maxMut={}_comp={}_permut={}_lambd={}_tolNMF={}.mat'
# .format(influence_weight, simplification, alpha, tol,
# keep_singletons, ngh_max,
# min_mutation, max_mutation,
# n_components, n_permutations, lambd, tol_nmf))
#
# consensus_data = loadmat(consensus_file)
# distance_genes = consensus_data['distance_genes']
# distance_patients = consensus_data['distance_patients']
hierarchical_clustering.distances_from_consensus_file(
result_folder, distance_genes, distance_patients, ppi_data, mut_type,
influence_weight, simplification,
alpha, tol, keep_singletons, ngh_max, min_mutation, max_mutation,
n_components, n_permutations, lambd, tol_nmf, linkage_method,
patient_data, data_folder, ssc_subgroups, ssc_mutation_data, gene_data)
(total_cluster_list, probands_cluster_list, siblings_cluster_list,
male_cluster_list, female_cluster_list, iq_cluster_list,
distCEU_list, mutation_nb_cluster_list,
text_file) = hierarchical_clustering.get_lists_from_clusters(
data_folder, patient_data, ssc_mutation_data,
ssc_subgroups, ppi_data, gene_data, result_folder,
mut_type, influence_weight, simplification, alpha, tol,
keep_singletons, ngh_max, min_mutation, max_mutation,
n_components, n_permutations, lambd, tol_nmf,
linkage_method)
hierarchical_clustering.bio_statistics(
n_components, total_cluster_list, probands_cluster_list,
siblings_cluster_list, male_cluster_list, female_cluster_list,
iq_cluster_list, distCEU_list, mutation_nb_cluster_list, text_file)
hierarchical_clustering.get_entrezgene_from_cluster(
data_folder, result_folder, ssc_mutation_data, patient_data,
ssc_subgroups, alpha, n_components, ngh_max, n_permutations, lambd,
influence_weight, simplification, tol, keep_singletons, min_mutation,
max_mutation, tol_nmf, linkage_method, gene_data, ppi_data,
gene_id_ppi, idx_ppi, idx_ppi_only, mut_type)
def all_functions(params):
if alpha == 0 and qn is not None:
print('############ PASS ############')
pass
else:
if patient_data == 'SSC':
result_folder = (data_folder + 'result_' + ssc_mutation_data +
'_' + ssc_subgroups + '_' + gene_data + '_' +
ppi_data + '/')
else:
result_folder = (data_folder + 'result_' + patient_data + '_' +
ppi_data + '/')
print(result_folder, flush=True)
print("alpha =", alpha, flush=True)
print("QN =", qn, flush=True)
print("k =", n_components, flush=True)
print("lambda =", lambd, flush=True)
print("PPI network =", ppi_data, flush=True)
# ------------ load_data.py ------------
print("------------ load_data.py ------------ {}".format(
datetime.now().strftime('%Y-%m-%d %H:%M:%S')))
if patient_data == 'TCGA_UCEC':
(patient_id, mutation_profile, gene_id_patient, gene_symbol_profile
) = load_data.load_TCGA_UCEC_patient_data(data_folder)
elif patient_data == 'Faroe':
mutation_profile, gene_id_patient = (
load_data.load_Faroe_Islands_data(data_folder))
elif patient_data == 'SSC':
mutation_profile, gene_id_patient, patient_id = (
load_data.load_specific_SSC_mutation_profile(
data_folder, ssc_mutation_data, ssc_subgroups, gene_data))
if ppi_data == 'Hofree_STRING':
gene_id_ppi, network = load_data.load_Hofree_PPI_String(
data_folder, ppi_data)
else:
gene_id_ppi, network = load_data.load_PPI_network(
data_folder, ppi_data)
# ------------ formatting_data.py ------------
print("------------ formatting_data.py ------------ {}".format(
datetime.now().strftime('%Y-%m-%d %H:%M:%S')))
(network, mutation_profile, idx_ppi, idx_mut, idx_ppi_only,
idx_mut_only) = (formatting_data.classify_gene_index(
network, mutation_profile, gene_id_ppi, gene_id_patient))
(ppi_total, mut_total, ppi_filt,
mut_filt) = (formatting_data.all_genes_in_submatrices(
network, idx_ppi, idx_mut, idx_ppi_only, idx_mut_only,
mutation_profile))
# ------------ filtering_diffusion.py ------------
print("------------ filtering_diffusion.py ------------ {}".format(
datetime.now().strftime('%Y-%m-%d %H:%M:%S')))
# ppi_influence = (
# filtering_diffusion.calcul_ppi_influence(
# sp.eye(ppi_filt.shape[0]), ppi_filt,
# result_folder, compute, overwrite, alpha, tol))
final_influence = (filtering_diffusion.calcul_final_influence(
sp.eye(ppi_filt.shape[0],
dtype=np.float32), ppi_filt, result_folder,
influence_weight, simplification, compute, overwrite, alpha, tol))
ppi_final, mut_final = filtering_diffusion.filter_ppi_patients(
ppi_total, mut_total, ppi_filt, final_influence, ngh_max,
keep_singletons, min_mutation, max_mutation)
mut_type, mut_propag = filtering_diffusion.propagation_profile(
mut_final, ppi_filt, result_folder, alpha, tol, qn)
# ------------ clustering.py ------------
print("------------ clustering.py ------------ {}".format(
datetime.now().strftime('%Y-%m-%d %H:%M:%S')))
genes_clustering, patients_clustering = (clustering.bootstrap(
result_folder, mut_type, mut_propag, ppi_final, influence_weight,
simplification, alpha, tol, keep_singletons, ngh_max, min_mutation,
max_mutation, n_components, n_permutations, run_bootstrap, lambd,
tol_nmf, compute_gene_clustering))
distance_genes, distance_patients = clustering.consensus_clustering(
result_folder, genes_clustering, patients_clustering,
influence_weight, simplification, mut_type, alpha, tol,
keep_singletons, ngh_max, min_mutation, max_mutation, n_components,
n_permutations, run_consensus, lambd, tol_nmf,
compute_gene_clustering)
# ------------ hierarchical_clustering.py ------------
print("------------ hierarchical_clustering.py ------------ {}".format(
datetime.now().strftime('%Y-%m-%d %H:%M:%S')))
hierarchical_clustering.distance_patients_from_consensus_file(
result_folder, distance_patients, ppi_data, mut_type,
influence_weight, simplification, alpha, tol, keep_singletons,
ngh_max, min_mutation, max_mutation, n_components, n_permutations,
lambd, tol_nmf, linkage_method, patient_data, data_folder,
ssc_subgroups, ssc_mutation_data, gene_data)
(total_cluster_list, probands_cluster_list, siblings_cluster_list,
male_cluster_list, female_cluster_list, iq_cluster_list, distCEU_list,
mutation_nb_cluster_list,
text_file) = hierarchical_clustering.get_lists_from_clusters(
data_folder, patient_data, ssc_mutation_data, ssc_subgroups,
ppi_data, gene_data, result_folder, mut_type, influence_weight,
simplification, alpha, tol, keep_singletons, ngh_max,
min_mutation, max_mutation, n_components, n_permutations, lambd,
tol_nmf, linkage_method)
hierarchical_clustering.bio_statistics(
n_components, total_cluster_list, probands_cluster_list,
siblings_cluster_list, male_cluster_list, female_cluster_list,
iq_cluster_list, distCEU_list, mutation_nb_cluster_list, text_file)
def merge_2_subgroups(data_folder, ssc_mutation_data, ssc_subgroups, gene_data,
ppi_data, n_components, mut_type, alpha, ngh_max,
n_permutations, lambd, O, gene2go, min_category_size,
max_category_size, max_category_depth):
################ SSC 1 ################
ssc_subgroups = 'SSC1'
print(ssc_subgroups)
mutation_profile, gene_id_patient, patient_id = (
load_data.load_specific_SSC_mutation_profile(
data_folder, ssc_mutation_data, ssc_subgroups, gene_data))
gene_id_ppi, network = load_data.load_PPI_network(data_folder, ppi_data)
network, mutation_profile, idx_ppi, idx_mut, idx_ppi_only, idx_mut_only = (
formatting_data.classify_gene_index(
network, mutation_profile, gene_id_ppi, gene_id_patient))
# idx_mut : List of common genes' indexes in patients' mutation profiles.
reference = pd.DataFrame({'GeneID': [gene_id_patient[i] for i in idx_mut]})
background = tools.generate_background(
gene2go, reference, 'GO_ID', 'GeneID')
goenrich.enrich.propagate(O, background, 'reference')
df_enrich1, df_pcount1 = p_df_by_subgroup(
data_folder, ssc_mutation_data, ssc_subgroups, gene_data, ppi_data,
n_components, mut_type, alpha, ngh_max, n_permutations, lambd, O,
min_category_size, max_category_size, max_category_depth)
################ SSC 2 ################
ssc_subgroups = 'SSC2'
print(ssc_subgroups)
mutation_profile, gene_id_patient, patient_id = (
load_data.load_specific_SSC_mutation_profile(
data_folder, ssc_mutation_data, ssc_subgroups, gene_data))
gene_id_ppi, network = load_data.load_PPI_network(data_folder, ppi_data)
network, mutation_profile, idx_ppi, idx_mut, idx_ppi_only, idx_mut_only = (
formatting_data.classify_gene_index(
network, mutation_profile, gene_id_ppi, gene_id_patient))
# idx_mut : List of common genes' indexes in patients' mutation profiles.
reference = pd.DataFrame({'GeneID': [gene_id_patient[i] for i in idx_mut]})
background = tools.generate_background(
gene2go, reference, 'GO_ID', 'GeneID')
goenrich.enrich.propagate(O, background, 'reference')
df_enrich2, df_pcount2 = p_df_by_subgroup(
data_folder, ssc_mutation_data, ssc_subgroups, gene_data, ppi_data,
n_components, mut_type, alpha, ngh_max, n_permutations, lambd, O,
min_category_size, max_category_size, max_category_depth)
################ merge ################
df_enrich = pd.concat([df_enrich1, df_enrich2], ignore_index=True)
# change col order
cols = list(df_enrich)
cols.insert(0, cols.pop(cols.index('k')))
cols.insert(0, cols.pop(cols.index('ind_group')))
df_enrich = df_enrich.loc[:, cols]
df_pcount = df_pcount1.merge(df_pcount2, how='inner', left_on='k',
right_on='k', suffixes=[1, 2])
directory = (
data_folder + 'result_biostat_genes_GOslim/' + ssc_mutation_data +
'_' + gene_data + '/')
os.makedirs(directory, exist_ok=True)
file_name_enrich = 'GOslim_{}_lambd={}_{}.pkl'.format(
mut_type, lambd, ppi_data)
file_name_pcount = 'p_count_{}_lambd={}_{}.pkl'.format(
mut_type, lambd, ppi_data)
# save
df_enrich.to_pickle('{}{}'.format(directory, file_name_enrich))
df_pcount.to_pickle('{}{}'.format(directory, file_name_pcount))
def all_functions(params):
# if patient_data == 'SSC':
# if mut_type == 'raw':
# alpha = 0
# result_folder = (
# data_folder + 'result_' + ssc_mutation_data + '_' +
# ssc_subgroups + '_' + gene_data + '/' + mut_type + '/')
# else:
# result_folder = (
# data_folder + 'result_' + ssc_mutation_data + '_' +
# ssc_subgroups + '_' + gene_data + '_' + ppi_data + '/')
# # result_folder = (
# # '/Volumes/Abu3/min/201809_sfari_without_category6_NaN/result_' + ssc_mutation_data + '_' +
# # ssc_subgroups + '_' + gene_data + '_' + ppi_data + '/')
# else:
# result_folder = (data_folder + 'result_' + patient_data + '_' +
# ppi_data + '/')
# if mut_type == 'raw':
# alpha = 0
# global params
muttype = params['muttype']
alpha, result_folder = initiation(mut_type, alpha, patient_data,
data_folder, ssc_mutation_data,
ssc_subgroups, gene_data, ppi_data,
lambd, n_components)
print(result_folder, flush=True)
print("mutation type =", mut_type, flush=True)
print("alpha =", alpha, flush=True)
print("k =", n_components, flush=True)
print("lambda =", lambd, flush=True)
print("PPI network =", ppi_data, flush=True)
# ------------ load_data.py ------------
print("------------ load_data.py ------------ {}".format(
datetime.now().strftime('%Y-%m-%d %H:%M:%S')))
if patient_data == 'TCGA_UCEC':
(patient_id, mutation_profile, gene_id_patient, gene_symbol_profile
) = load_data.load_TCGA_UCEC_patient_data(data_folder)
elif patient_data == 'Faroe':
mutation_profile, gene_id_patient = (
load_data.load_Faroe_Islands_data(data_folder))
elif patient_data == 'SSC':
mutation_profile, gene_id_patient, patient_id = (
load_data.load_specific_SSC_mutation_profile(
data_folder, ssc_mutation_data, ssc_subgroups, gene_data))
if ppi_data == 'Hofree_STRING':
gene_id_ppi, network = load_data.load_Hofree_PPI_String(
data_folder, ppi_data)
else:
gene_id_ppi, network = load_data.load_PPI_network(
data_folder, ppi_data)
# ------------ formatting_data.py ------------
print("------------ formatting_data.py ------------ {}".format(
datetime.now().strftime('%Y-%m-%d %H:%M:%S')))
(network, mutation_profile, idx_ppi, idx_mut, idx_ppi_only,
idx_mut_only) = (formatting_data.classify_gene_index(
network, mutation_profile, gene_id_ppi, gene_id_patient))
(ppi_total, mut_total, ppi_filt,
mut_filt) = (formatting_data.all_genes_in_submatrices(
network, idx_ppi, idx_mut, idx_ppi_only, idx_mut_only,
mutation_profile))
# ------------ filtering_diffusion.py ------------
print("------------ filtering_diffusion.py ------------ {}".format(
datetime.now().strftime('%Y-%m-%d %H:%M:%S')))
# ppi_influence = (
# filtering_diffusion.calcul_ppi_influence(
# sp.eye(ppi_filt.shape[0]), ppi_filt,
# result_folder, compute, overwrite, alpha, tol))
final_influence = (filtering_diffusion.calcul_final_influence(
sp.eye(ppi_filt.shape[0], dtype=np.float32), ppi_filt, result_folder,
influence_weight, simplification, compute, overwrite, alpha, tol))
ppi_final, mut_final = filtering_diffusion.filter_ppi_patients(
ppi_total, mut_total, ppi_filt, final_influence, ngh_max,
keep_singletons, min_mutation, max_mutation)
mut_type, mut_propag = filtering_diffusion.propagation_profile(
mut_final, ppi_filt, result_folder, alpha, tol, qn)
# ------------ clustering.py ------------
print("------------ clustering.py ------------ {}".format(
datetime.now().strftime('%Y-%m-%d %H:%M:%S')))
genes_clustering, patients_clustering = (clustering.bootstrap(
result_folder, mut_type, mut_propag, ppi_final, influence_weight,
simplification, alpha, tol, keep_singletons, ngh_max, min_mutation,
max_mutation, n_components, n_permutations, run_bootstrap, lambd,
tol_nmf, compute_gene_clustering))
distance_genes, distance_patients = clustering.consensus_clustering(
result_folder, genes_clustering, patients_clustering, influence_weight,
simplification, mut_type, alpha, tol, keep_singletons, ngh_max,
min_mutation, max_mutation, n_components, n_permutations,
run_consensus, lambd, tol_nmf, compute_gene_clustering)
# ------------ hierarchical_clustering.py ------------
print("------------ hierarchical_clustering.py ------------ {}".format(
datetime.now().strftime('%Y-%m-%d %H:%M:%S')))
hierarchical_clustering.distance_patients_from_consensus_file(
result_folder, distance_patients, ppi_data, mut_type, influence_weight,
simplification, alpha, tol, keep_singletons, ngh_max, min_mutation,
max_mutation, n_components, n_permutations, lambd, tol_nmf,
linkage_method, patient_data, data_folder, ssc_subgroups,
ssc_mutation_data, gene_data)
(total_cluster_list, probands_cluster_list, siblings_cluster_list,
male_cluster_list, female_cluster_list, iq_cluster_list, distCEU_list,
mutation_nb_cluster_list,
text_file) = hierarchical_clustering.get_lists_from_clusters(
data_folder, patient_data, ssc_mutation_data, ssc_subgroups, ppi_data,
gene_data, result_folder, mut_type, influence_weight, simplification,
alpha, tol, keep_singletons, ngh_max, min_mutation, max_mutation,
n_components, n_permutations, lambd, tol_nmf, linkage_method)
hierarchical_clustering.bio_statistics(
n_components, total_cluster_list, probands_cluster_list,
siblings_cluster_list, male_cluster_list, female_cluster_list,
iq_cluster_list, distCEU_list, mutation_nb_cluster_list, text_file)
print("\n------------ biostat.py ------------ {}".format(
datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S')),
flush=True)
import biostat
gene_id_ppi, idx_ppi, idx_ppi_only = preprocessing(
data_folder, patient_data, ssc_mutation_data, ssc_subgroups, gene_data,
ppi_data, result_folder, influence_weight, simplification, compute,
overwrite, alpha, tol, ngh_max, keep_singletons, min_mutation,
max_mutation, mut_type)
biostat.biostat_analysis(
data_folder, result_folder, patient_data, ssc_mutation_data,
ssc_subgroups, ppi_data, gene_data, mut_type, influence_weight,
simplification, alpha, tol, keep_singletons, ngh_max, min_mutation,
max_mutation, n_components, n_permutations, lambd, tol_nmf,
linkage_method, p_val_threshold, gene_id_ppi, idx_ppi, idx_ppi_only)
biostat_go.biostat_go_enrichment(alpha, result_folder, mut_type,
patient_data, data_folder,
ssc_mutation_data, ssc_subgroups,
gene_data, ppi_data, lambd, n_components,
ngh_max, n_permutations)
