def run_gene_prioritization_pipeline(self):
"""
Runs data cleaning for gene_prioritization_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.
"""
# Checks user spreadsheet data and phenotype data
if self.user_spreadsheet_df is None or self.phenotype_df is None:
return False, logger.logging
# Imputes na value on user spreadsheet data
user_spreadsheet_df_imputed = SpreadSheet.impute_na(self.user_spreadsheet_df,
option=self.run_parameters['impute'])
if user_spreadsheet_df_imputed is None:
return False, logger.logging
# Checks if value of inputs satisfy certain criteria: see details in function validate_inputs_for_gp_fp
user_spreadsheet_val_chked, phenotype_val_checked = CommonUtil.validate_inputs_for_gp_fp(
user_spreadsheet_df_imputed, self.phenotype_df, self.run_parameters['correlation_measure'])
if user_spreadsheet_val_chked is None or phenotype_val_checked 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 or phenotype_val_checked is None:
return False, logger.logging
# Stores cleaned phenotype data (transposed) to a file, dimension: phenotype x sample
IOUtil.write_to_file(phenotype_val_checked, 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')
# 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]))
logger.logging.append(
'INFO: Cleaned phenotype data has {} row(s), {} column(s).'.format(phenotype_val_checked.shape[0],
phenotype_val_checked.shape[1]))
return True, logger.logging
def run_feature_prioritization_pipeline(self):
"""
Run data cleaning for feature prioritization 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.
"""
from knpackage.toolbox import get_spreadsheet_df
if self.user_spreadsheet_df is None or self.phenotype_df is None:
return False, logger.logging
# Imputes na value on user spreadsheet data
user_spreadsheet_df_imputed = SpreadSheet.impute_na(self.user_spreadsheet_df,
option=self.run_parameters['impute'])
if user_spreadsheet_df_imputed is None:
return False, logger.logging
# Checks if value of inputs satisfy certain criteria
user_spreadsheet_val_chked, phenotype_val_chked = CommonUtil.validate_inputs_for_gp_fp(
user_spreadsheet_df_imputed,
self.phenotype_df, self.run_parameters[
'correlation_measure'])
if user_spreadsheet_val_chked is None or phenotype_val_chked is None:
return False, logger.logging
IOUtil.write_to_file(user_spreadsheet_val_chked, 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_val_chked.shape[0],
user_spreadsheet_val_chked.shape[1]))
if self.run_parameters['correlation_measure'] == 't_test':
phenotype_df = get_spreadsheet_df(self.run_parameters['phenotype_name_full_path'])
phenotype_output = TransformationUtil.phenotype_expander(phenotype_df)
else:
phenotype_output = phenotype_val_chked
IOUtil.write_to_file(phenotype_output, self.run_parameters['phenotype_name_full_path'],
self.run_parameters['results_directory'], '_ETL.tsv')
logger.logging.append(
'INFO: Cleaned phenotypic data has {} row(s), {} column(s).'.format(phenotype_val_chked.shape[0],
phenotype_val_chked.shape[1]))
return True, logger.logging