p3/preprocess/Ovarian_Cancer/code/GSE146553.py

# Path Configuration
from tools.preprocess import *

# Processing context
trait = "Ovarian_Cancer"
cohort = "GSE146553"

# Input paths
in_trait_dir = "../DATA/GEO/Ovarian_Cancer"
in_cohort_dir = "../DATA/GEO/Ovarian_Cancer/GSE146553"

# Output paths
out_data_file = "./output/preprocess/3/Ovarian_Cancer/GSE146553.csv"
out_gene_data_file = "./output/preprocess/3/Ovarian_Cancer/gene_data/GSE146553.csv"
out_clinical_data_file = "./output/preprocess/3/Ovarian_Cancer/clinical_data/GSE146553.csv"
json_path = "./output/preprocess/3/Ovarian_Cancer/cohort_info.json"

# Get file paths for SOFT and matrix files
soft_file_path, matrix_file_path = geo_get_relevant_filepaths(in_cohort_dir)

# Get background info and clinical data from the matrix file
background_info, clinical_data = get_background_and_clinical_data(matrix_file_path)

# Create dictionary of unique values for each feature
unique_values_dict = get_unique_values_by_row(clinical_data)

# Print the information
print("Dataset Background Information:")
print(background_info)
print("\nSample Characteristics:")
for feature, values in unique_values_dict.items():
    print(f"\n{feature}:")
    print(values)
# 1. Gene Expression Data Availability
# From the background info, this dataset contains Affymetrix gene expression data
is_gene_available = True

# 2. Variable Availability and Data Type Conversion
# 2.1 Data Availability 
# Trait (cancer status) can be inferred from tissue type in row 4
trait_row = 4

# Age is available in row 2
age_row = 2 

# Gender is available in row 5 but shows only one value 'female'
# Constant feature is not useful, so set to None
gender_row = None

# 2.2 Data Type Conversion
def convert_trait(value):
    if not isinstance(value, str):
        return None
    # Extract value after colon
    val = value.split(': ')[-1].lower()
    # Normal tissue = 0, tumor tissue = 1
    if 'normal' in val:
        return 0
    elif 'cancer' in val or 'tumor' in val:
        return 1
    else:
        return None

def convert_age(value):
    if not isinstance(value, str):
        return None
    try:
        # Extract numeric value after colon
        age = float(value.split(': ')[-1])
        return age
    except:
        return None

def convert_gender(value):
    # Not used since gender is constant
    pass

# 3. Save Metadata 
# Initial filtering - only checking data availability
validate_and_save_cohort_info(
    is_final=False,
    cohort=cohort,
    info_path=json_path,
    is_gene_available=is_gene_available,
    is_trait_available=(trait_row is not None)
)

# 4. Clinical Feature Extraction
# Since trait_row is not None, extract clinical features
clinical_df = geo_select_clinical_features(
    clinical_data,
    trait=trait,
    trait_row=trait_row,
    convert_trait=convert_trait,
    age_row=age_row,
    convert_age=convert_age
)

# Preview the extracted features
print("Preview of clinical features:")
print(preview_df(clinical_df))

# Save clinical data
clinical_df.to_csv(out_clinical_data_file)
# Extract genetic data matrix
genetic_data = get_genetic_data(matrix_file_path)

# Print first few rows with column names to examine data structure
print("Data preview:")
print("\nColumn names:")
print(list(genetic_data.columns)[:5])
print("\nFirst 5 rows:")
print(genetic_data.head())
print("\nShape:", genetic_data.shape)

# Verify this is gene expression data and check identifiers
is_gene_available = True

# Save updated metadata
validate_and_save_cohort_info(
    is_final=False,
    cohort=cohort,
    info_path=json_path, 
    is_gene_available=is_gene_available,
    is_trait_available=(trait_row is not None)
)

# Save gene expression data 
genetic_data.to_csv(out_gene_data_file)
# The row IDs are numerical Illumina BeadChip identifiers (like 7896736, 7896738), not gene symbols
# These need to be mapped to proper human gene symbols for analysis
requires_gene_mapping = True
# Extract gene annotation data
gene_metadata = get_gene_annotation(soft_file_path) 

# Preview column names and first few values
preview = preview_df(gene_metadata)
print("\nGene annotation columns and sample values:")
print(preview)
# 1. From preview, we can see 'ID' column contains probe IDs matching those in gene expression data,
# and 'gene_assignment' column contains gene symbols in format "RefSeq // SYMBOL // Description"

# 2. Extract probe-gene pairs from gene annotation data
def extract_gene_symbol(text):
    if not isinstance(text, str):
        return None
    # Split by // and extract the second field which contains the gene symbol
    fields = text.split('//')
    if len(fields) >= 2:
        return fields[1].strip()
    return None

# Copy gene_metadata and add parsed gene symbols column
gene_metadata_with_symbols = gene_metadata.copy()
gene_metadata_with_symbols['Gene'] = gene_metadata_with_symbols['gene_assignment'].apply(extract_gene_symbol)

# Get mapping between probe IDs and gene symbols
mapping_df = get_gene_mapping(gene_metadata_with_symbols, prob_col='ID', gene_col='Gene')

# 3. Convert probe-level data to gene expression using many-to-many mapping
gene_data = apply_gene_mapping(genetic_data, mapping_df)

# Preview the results
print("\nShape before mapping:", genetic_data.shape)
print("Shape after mapping:", gene_data.shape)
print("\nFirst few gene symbols:")
print(list(gene_data.index)[:10])

# Save gene expression data after mapping
gene_data.to_csv(out_gene_data_file)
# 1. Normalize gene symbols and save gene data
normalized_gene_data = normalize_gene_symbols_in_index(gene_data)
os.makedirs(os.path.dirname(out_gene_data_file), exist_ok=True)
normalized_gene_data.to_csv(out_gene_data_file)

# 2. Link clinical and genetic data  
clinical_features = pd.read_csv(out_clinical_data_file, index_col=0)
linked_data = geo_link_clinical_genetic_data(clinical_features, normalized_gene_data)

# 3. Handle missing values
linked_data = handle_missing_values(linked_data, trait)

# 4. Judge bias in features and remove biased ones
trait_biased, linked_data = judge_and_remove_biased_features(linked_data, trait)

# 5. Final validation and save metadata
is_usable = validate_and_save_cohort_info(
    is_final=True,
    cohort=cohort,
    info_path=json_path,
    is_gene_available=is_gene_available, 
    is_trait_available=True,
    is_biased=trait_biased,
    df=linked_data,
    note="Gene expression data from melanoma patients receiving PD-1 immunotherapy, with long-term benefit as outcome."
)

# 6. Save linked data if usable
if is_usable:
    os.makedirs(os.path.dirname(out_data_file), exist_ok=True)
    linked_data.to_csv(out_data_file)