def run_general_clustering_pipeline(self):
"""
Runs data cleaning for general_clustering_pipeline.
Args:
NA.
Returns:
validation_flag: Boolean type value indicating if input data is valid or not.
message: A message indicates the status of current check.
"""
if self.user_spreadsheet_df is None:
return False, logger.logging
# Checks intersection between user spreadsheet data and phenotype data
phenotype_df_cleaned = None
if self.phenotype_df is not None:
phenotype_df_cleaned = CommonUtil.check_phenotype_intersection(self.phenotype_df,
self.user_spreadsheet_df.columns.values)
if phenotype_df_cleaned is None:
logger.logging.append('ERROR: Phenotype is emtpy. Please provide a valid phenotype data.')
return False, logger.logging
logger.logging.append('INFO: Start to process user spreadsheet data.')
# Checks if user spreadsheet contains na value and only real number
user_spreadsheet_df_val_check = SpreadSheet.check_user_spreadsheet_data(self.user_spreadsheet_df,
dropna_colwise=True,
check_real_number=True,
check_positive_number=True)
if user_spreadsheet_df_val_check is None:
return False, logger.logging
user_spreadsheet_df_rm_na_header = SpreadSheet.remove_na_header(user_spreadsheet_df_val_check)
if user_spreadsheet_df_rm_na_header is None:
return False, logger.logging
# Removes NA value and duplication on column and row name
user_spreadsheet_df_cleaned = SpreadSheet.remove_dataframe_indexer_duplication(user_spreadsheet_df_rm_na_header)
if user_spreadsheet_df_cleaned is None:
return False, logger.logging
IOUtil.write_to_file(user_spreadsheet_df_cleaned, self.run_parameters['spreadsheet_name_full_path'],
self.run_parameters['results_directory'], '_ETL.tsv')
logger.logging.append(
'INFO: Cleaned user spreadsheet has {} row(s), {} column(s).'.format(
user_spreadsheet_df_cleaned.shape[0],
user_spreadsheet_df_cleaned.shape[1]))
if phenotype_df_cleaned is not None:
IOUtil.write_to_file(phenotype_df_cleaned, self.run_parameters['phenotype_name_full_path'],
self.run_parameters['results_directory'], '_ETL.tsv')
logger.logging.append(
'INFO: Cleaned phenotype data has {} row(s), {} column(s).'.format(phenotype_df_cleaned.shape[0],
phenotype_df_cleaned.shape[1]))
return True, logger.logging
def run_geneset_characterization_pipeline(self):
"""
Runs data cleaning for geneset_characterization_pipeline.
Args:
NA
Returns:
validation_flag: Boolean type value indicating if input data is valid or not.
message: A message indicates the status of current check.
"""
if self.user_spreadsheet_df is None:
return False, logger.logging
# Checks only non-negative real number appears in user spreadsheet, drop na column wise
user_spreadsheet_val_chked = SpreadSheet.check_user_spreadsheet_data(self.user_spreadsheet_df, check_na=True,
check_real_number=True,
check_positive_number=True)
if user_spreadsheet_val_chked is None:
return False, logger.logging
# Removes NA value and duplication on column and row name
user_spreadsheet_df_checked = SpreadSheet.remove_dataframe_indexer_duplication(user_spreadsheet_val_chked)
# Checks the validity of gene name to see if it can be ensemble or not
user_spreadsheet_df_cleaned, map_filtered_dedup, mapping = SpreadSheet.map_ensemble_gene_name(
user_spreadsheet_df_checked,
self.run_parameters)
if user_spreadsheet_df_cleaned is None:
return False, logger.logging
IOUtil.write_to_file(user_spreadsheet_df_cleaned, self.run_parameters['spreadsheet_name_full_path'],
self.run_parameters['results_directory'], '_ETL.tsv')
# writes dedupped mapping between user_supplied_gene_name and ensemble name to a file
IOUtil.write_to_file(map_filtered_dedup, self.run_parameters['spreadsheet_name_full_path'],
self.run_parameters['results_directory'], '_MAP.tsv', use_index=True, use_header=False)
# writes user supplied gene name along with its mapping status to a file
IOUtil.write_to_file(mapping, self.run_parameters['spreadsheet_name_full_path'],
self.run_parameters['results_directory'],
'_User_To_Ensembl.tsv', use_index=False, use_header=True)
logger.logging.append(
'INFO: Cleaned user spreadsheet has {} row(s), {} column(s).'.format(
user_spreadsheet_df_cleaned.shape[0],
user_spreadsheet_df_cleaned.shape[1]))
return True, logger.logging
def run_phenotype_prediction_pipeline(self):
"""
Runs data cleaning for phenotype_prediction_pipeline.
Args:
NA.
Returns:
validation_flag: Boolean type value indicating if input data is valid or not.
message: A message indicates the status of current check.
"""
# spreadsheet dimension: sample x phenotype, phenotype dimension : sample x phenotype
if self.user_spreadsheet_df is None or self.phenotype_df is None:
return False, logger.logging
# Checks if user spreadsheet contains only real number and drop na column wise
user_spreadsheet_dropna = SpreadSheet.check_user_spreadsheet_data(self.user_spreadsheet_df,
dropna_colwise=True,
check_real_number=True)
if user_spreadsheet_dropna is None or user_spreadsheet_dropna.empty:
logger.logging.append('ERROR: After drop NA, user spreadsheet data becomes empty.')
return None, None
# Checks for valid intersection between phenotype data and user spreadsheet data
dataframe_header = list(user_spreadsheet_dropna.columns.values)
phenotype_df_pxs_trimmed = CheckUtil.check_intersection_for_phenotype_and_user_spreadsheet(dataframe_header,
self.phenotype_df)
# Removes NA value and duplication on column and row name
user_spreadsheet_df_cleaned = SpreadSheet.remove_dataframe_indexer_duplication(user_spreadsheet_dropna)
if user_spreadsheet_df_cleaned is None or phenotype_df_pxs_trimmed is None:
return False, logger.logging
# Stores cleaned phenotype data (transposed) to a file, dimension: phenotype x sample
IOUtil.write_to_file(phenotype_df_pxs_trimmed, self.run_parameters['phenotype_name_full_path'],
self.run_parameters['results_directory'], '_ETL.tsv')
IOUtil.write_to_file(user_spreadsheet_df_cleaned, self.run_parameters['spreadsheet_name_full_path'],
self.run_parameters['results_directory'], '_ETL.tsv')
logger.logging.append(
'INFO: Cleaned user spreadsheet has {} row(s), {} column(s).'.format(
user_spreadsheet_df_cleaned.shape[0],
user_spreadsheet_df_cleaned.shape[1]))
logger.logging.append(
'INFO: Cleaned phenotype data has {} row(s), {} column(s).'.format(phenotype_df_pxs_trimmed.shape[0],
phenotype_df_pxs_trimmed.shape[1]))
return True, logger.logging
def run_simplified_inpherno_pipeline(self):
"""
Runs data cleaning for simplified_inpherno_pipeline.
Args:
NA.
Returns:
validation_flag: Boolean type value indicating if input data is valid or not.
message: A message indicates the status of current check.
"""
output_files = ['Pvalue_gene_phenotype', 'expression_sample', 'TFexpression']
for file in output_files:
if eval(str('self.' + file)) is None:
return False, logger.logging
for file in output_files:
cur_data = eval(str('self.' + file))
if SpreadSheet.check_user_spreadsheet_data(cur_data, check_real_number=True,
check_na=True if file is 'TFexpression' else False) is None:
return False, logger.logging
cur_data_cleaned, mapping_dedup, mapping = SpreadSheet.map_ensemble_gene_name(cur_data, self.run_parameters)
if cur_data_cleaned is None:
return False, logger.logging
IOUtil.write_to_file(cur_data, self.run_parameters[file + '_full_path'],
self.run_parameters['results_directory'], '.tsv',
use_header=False if file is 'TFexpression' else True)
IOUtil.write_to_file(cur_data_cleaned, self.run_parameters[file + '_full_path'],
self.run_parameters['results_directory'], '_ETL.tsv',
use_header=False if file is 'TFexpression' else True)
# writes dedupped mapping between user_supplied_gene_name and ensemble name to a file
IOUtil.write_to_file(mapping_dedup, self.run_parameters[file + '_full_path'],
self.run_parameters['results_directory'], '_MAP.tsv', use_index=True, use_header=False)
# writes user supplied gene name along with its mapping status to a file
IOUtil.write_to_file(mapping, self.run_parameters[file + '_full_path'],
self.run_parameters['results_directory'],
'_User_To_Ensembl.tsv', use_index=False, use_header=True)
return True, logger.logging
def run_samples_clustering_pipeline(self):
"""
Runs data cleaning for samples_clustering_pipeline.
Args:
NA
Returns:
validation_flag: Boolean type value indicating if input data is valid or not.
message: A message indicates the status of current check.
"""
if self.user_spreadsheet_df is None:
return False, logger.logging
logger.logging.append('INFO: Start to process user spreadsheet data.')
# Checks if only non-negative real number appears in user spreadsheet and drop na column wise
user_spreadsheet_val_chked = SpreadSheet.check_user_spreadsheet_data(self.user_spreadsheet_df,
dropna_colwise=True,
check_real_number=True,
check_positive_number=True)
if user_spreadsheet_val_chked is None:
return False, logger.logging
# Removes NA value and duplication on column and row name
user_spreadsheet_df_checked = SpreadSheet.remove_dataframe_indexer_duplication(user_spreadsheet_val_chked)
# Checks the validity of gene name to see if it can be ensemble or not
user_spreadsheet_df_cleaned, map_filtered_dedup, mapping = SpreadSheet.map_ensemble_gene_name(
user_spreadsheet_df_checked,
self.run_parameters)
if 'gg_network_name_full_path' in self.run_parameters.keys() and \
not CommonUtil.check_network_data_intersection(user_spreadsheet_df_cleaned.index,
self.run_parameters):
return False, logger.logging
# The logic here ensures that even if phenotype data doesn't fits requirement, the rest pipelines can still run.
if user_spreadsheet_df_cleaned is None:
return False, logger.logging
else:
IOUtil.write_to_file(user_spreadsheet_df_cleaned, self.run_parameters['spreadsheet_name_full_path'],
self.run_parameters['results_directory'], '_ETL.tsv')
# writes dedupped mapping between user_supplied_gene_name and ensemble name to a file
IOUtil.write_to_file(map_filtered_dedup, self.run_parameters['spreadsheet_name_full_path'],
self.run_parameters['results_directory'], '_MAP.tsv', use_index=True, use_header=False)
# writes user supplied gene name along with its mapping status to a file
IOUtil.write_to_file(mapping, self.run_parameters['spreadsheet_name_full_path'],
self.run_parameters['results_directory'],
'_User_To_Ensembl.tsv', use_index=False, use_header=True)
logger.logging.append(
'INFO: Cleaned user spreadsheet has {} row(s), {} column(s).'.format(
user_spreadsheet_df_cleaned.shape[0],
user_spreadsheet_df_cleaned.shape[1]))
if self.phenotype_df is not None:
logger.logging.append('INFO: Start to process phenotype data.')
phenotype_df_cleaned = CommonUtil.check_phenotype_intersection(self.phenotype_df,
self.user_spreadsheet_df.columns.values)
if phenotype_df_cleaned is None:
logger.logging.append('ERROR: Phenotype is emtpy. Please provide a valid phenotype data.')
return False, logger.logging
else:
IOUtil.write_to_file(phenotype_df_cleaned, self.run_parameters['phenotype_name_full_path'],
self.run_parameters['results_directory'], '_ETL.tsv')
logger.logging.append('INFO: Cleaned phenotype data has {} row(s), {} '
'column(s).'.format(phenotype_df_cleaned.shape[0], phenotype_df_cleaned.shape[1]))
return True, logger.logging
def run_signature_analysis_pipeline(self):
"""
Runs data cleaning for signature_analysis_pipeline.
Args:
NA.
Returns:
validation_flag: Boolean type value indicating if input data is valid or not.
message: A message indicates the status of current check.
"""
if self.signature_df is None or self.user_spreadsheet_df is None:
return False, logger.logging
# Removes NA index for both signature data and user spreadsheet data
signature_df = SpreadSheet.remove_na_index(self.signature_df)
user_spreadsheet_df = SpreadSheet.remove_na_index(self.user_spreadsheet_df)
# Checks if only real number and non-NA value appear in user spreadsheet
if SpreadSheet.check_user_spreadsheet_data(user_spreadsheet_df, check_na=True,
check_real_number=True,
check_positive_number=False) is None:
return False, logger.logging
# Checks duplicate columns and rows in user spreadsheet data
if CheckUtil.check_duplicates(user_spreadsheet_df, check_column=True, check_row=True):
logger.logging.append('ERROR: Found duplicates on user spreadsheet data. Rejecting...')
return False, logger.logging
# Checks intersection of genes between signature data and user spreadsheet data
intersection = CheckUtil.find_intersection(signature_df.index, user_spreadsheet_df.index)
if intersection is None:
logger.logging.append('ERROR: Cannot find intersection between spreadsheet genes and signature genes.')
return False, logger.logging
logger.logging.append(
'INFO: Found {} intersected gene(s) between phenotype and spreadsheet data.'.format(len(intersection)))
# Checks number of unique value in userspread sheet equals to 2
if not SpreadSheet.check_unique_values(user_spreadsheet_df, cnt=2):
logger.logging.append(
'ERROR: user spreadsheet data does not meet the requirment of having at least two unique values.')
return False, logger.logging
# Checks intersection among network data, signature data and user spreadsheet data
if 'gg_network_name_full_path' in self.run_parameters.keys() and \
not CommonUtil.check_network_data_intersection(intersection,
self.run_parameters):
return False, logger.logging
# The logic here ensures that even if phenotype data doesn't fits requirement, the rest pipelines can still run.
if user_spreadsheet_df is None:
return False, logger.logging
else:
IOUtil.write_to_file(user_spreadsheet_df,
self.run_parameters['spreadsheet_name_full_path'],
self.run_parameters['results_directory'], '_ETL.tsv')
logger.logging.append(
'INFO: Cleaned user spreadsheet has {} row(s), {} column(s).'.format(
user_spreadsheet_df.shape[0],
user_spreadsheet_df.shape[1]))
if signature_df is not None:
IOUtil.write_to_file(signature_df, self.run_parameters['signature_name_full_path'],
self.run_parameters['results_directory'], '_ETL.tsv')
logger.logging.append(
'INFO: Cleaned phenotype data has {} row(s), {} column(s).'.format(signature_df.shape[0],
signature_df.shape[1]))
return True, logger.logging