p1/preprocess/Asthma/code/GSE123086.py

# Path Configuration
from tools.preprocess import *

# Processing context
trait = "Asthma"
cohort = "GSE123086"

# Input paths
in_trait_dir = "../DATA/GEO/Asthma"
in_cohort_dir = "../DATA/GEO/Asthma/GSE123086"

# Output paths
out_data_file = "./output/preprocess/1/Asthma/GSE123086.csv"
out_gene_data_file = "./output/preprocess/1/Asthma/gene_data/GSE123086.csv"
out_clinical_data_file = "./output/preprocess/1/Asthma/clinical_data/GSE123086.csv"
json_path = "./output/preprocess/1/Asthma/cohort_info.json"

# STEP1
from tools.preprocess import *
# 1. Identify the paths to the SOFT file and the matrix file
soft_file, matrix_file = geo_get_relevant_filepaths(in_cohort_dir)

# 2. Read the matrix file to obtain background information and sample characteristics data
background_prefixes = ['!Series_title', '!Series_summary', '!Series_overall_design']
clinical_prefixes = ['!Sample_geo_accession', '!Sample_characteristics_ch1']
background_info, clinical_data = get_background_and_clinical_data(matrix_file, background_prefixes, clinical_prefixes)

# 3. Obtain the sample characteristics dictionary from the clinical dataframe
sample_characteristics_dict = get_unique_values_by_row(clinical_data)

# 4. Explicitly print out all the background information and the sample characteristics dictionary
print("Background Information:")
print(background_info)
print("Sample Characteristics Dictionary:")
print(sample_characteristics_dict)
# 1. Determine if the dataset likely contains gene expression data
#    Based on the microarray-based gene expression description, set this to True.
is_gene_available = True

# 2. Identify availability of "trait", "age", and "gender" from the sample characteristics
#    After examining each row in the sample characteristics dictionary:
#    - The primary diagnosis row is key 1, which includes "primary diagnosis: ASTHMA" among others.
#      That will serve as our trait row, since it's not constant and contains "ASTHMA".
#    - Age values appear predominantly in row 3 (and some in row 4). We'll select row 3 for age.
#    - Gender data is scattered (partly in row 2, partly in row 3) and not presented in a single row,
#      so we set gender_row to None.

trait_row = 1
age_row = 3
gender_row = None

# 2.2. Define data conversion functions

def convert_trait(x: str) -> Optional[int]:
    """
    Convert trait data into a binary variable, 1 for ASTHMA, 0 otherwise. 
    If not parsable, return None.
    """
    parts = x.split(':')
    if len(parts) < 2:
        return None
    val = parts[1].strip().upper()
    return 1 if val == "ASTHMA" else 0

def convert_age(x: str) -> Optional[float]:
    """
    Convert age data into a continuous float. If the string does not 
    contain 'age:' or cannot be parsed, return None.
    """
    parts = x.split(':')
    if len(parts) < 2:
        return None
    if "age" in parts[0].lower():
        try:
            return float(parts[1].strip())
        except ValueError:
            return None
    return None

def convert_gender(x: str) -> Optional[int]:
    """
    Convert gender data to 0 (female) or 1 (male). If not parsable, return None.
    """
    parts = x.split(':')
    if len(parts) < 2:
        return None
    if "sex" in parts[0].lower():
        val = parts[1].strip().lower()
        if val == "female":
            return 0
        elif val == "male":
            return 1
    return None

# 3. Save metadata (initial filtering)
#    Trait availability is True if trait_row is not None, otherwise False.
is_trait_available = (trait_row is not None)
is_usable = validate_and_save_cohort_info(
    is_final=False,
    cohort=cohort,
    info_path=json_path,
    is_gene_available=is_gene_available,
    is_trait_available=is_trait_available
)

# 4. If trait data is available, extract clinical features and save them
if trait_row is not None:
    selected_clinical_df = geo_select_clinical_features(
        clinical_df=clinical_data,
        trait=trait,
        trait_row=trait_row,
        convert_trait=convert_trait,
        age_row=age_row,
        convert_age=convert_age,
        gender_row=gender_row,
        convert_gender=convert_gender
    )
    # Preview the resulting DataFrame
    preview_clin = preview_df(selected_clinical_df)
    print("Preview of selected clinical features:", preview_clin)

    # Save the clinical data to CSV
    selected_clinical_df.to_csv(out_clinical_data_file, index=False)
# STEP3
# 1. Use the get_genetic_data function from the library to get the gene_data from the matrix_file previously defined.
gene_data = get_genetic_data(matrix_file)

# 2. Print the first 20 row IDs (gene or probe identifiers) for future observation.
print(gene_data.index[:20])
# Observing the identifiers: they appear to be numeric and not standard human gene symbols.
# Therefore, they likely need to be mapped to gene symbols.
print("requires_gene_mapping = True")
# STEP5
# 1. Use the 'get_gene_annotation' function from the library to get gene annotation data from the SOFT file.
gene_annotation = get_gene_annotation(soft_file)

# 2. Use the 'preview_df' function from the library to preview the data and print out the results.
print("Gene annotation preview:")
print(preview_df(gene_annotation))
# STEP 6: Gene Identifier Mapping (Revised Debugged Code)

def apply_gene_mapping_entrez(expression_df: pd.DataFrame, mapping_df: pd.DataFrame) -> pd.DataFrame:
    """
    Convert measured data about gene probes (indexed by numeric 'ID') into gene-level data
    (using numeric Entrez IDs). Handles one-to-many or many-to-one mappings by splitting
    probe expression values equally among mapped genes, and summing where multiple probes
    map to the same gene.
    """
    # Remove any duplicate probe entries in the mapping
    mapping_df = mapping_df.drop_duplicates(subset=['ID', 'Gene'])
    mapping_df = mapping_df.dropna(subset=['ID', 'Gene'])

    # Also ensure expression_df has a unique index
    expression_df = expression_df[~expression_df.index.duplicated(keep='first')]

    # Make sure mapping DataFrame is indexed by probe ID
    mapping_df.set_index('ID', inplace=True)

    # Some platforms may have multiple Entrez IDs joined by a delimiter. Split safely if needed.
    mapping_df['Gene'] = mapping_df['Gene'].astype(str)
    mapping_df['Gene'] = mapping_df['Gene'].apply(
        lambda x: x.split('//') if '//' in x else x.split(';') if ';' in x else [x]
    )

    # Count the number of genes each probe maps to
    mapping_df['num_genes'] = mapping_df['Gene'].apply(len)

    # Expand to one row per (probe, gene) pair
    mapping_df = mapping_df.explode('Gene').dropna(subset=['Gene'])

    # Join expression values (probe-level) onto the mapping table
    merged_df = mapping_df.join(expression_df, how='inner')  # inner join to keep only matched probes

    # Identify the columns containing actual expression values (the sample columns)
    # We'll exclude 'Gene' and 'num_genes'
    expr_cols = [c for c in merged_df.columns if c not in ['Gene', 'num_genes']]

    # Divide each probe's expression by the number of genes it maps to
    merged_df[expr_cols] = merged_df[expr_cols].div(merged_df['num_genes'].replace(0, 1), axis=0)

    # Finally, sum over genes to get gene-level expression data
    gene_expression_df = merged_df.groupby('Gene')[expr_cols].sum()

    return gene_expression_df

# 1. Identify the columns in the annotation that match our needs
probe_col = "ID"
gene_col = "ENTREZ_GENE_ID"

# 2. Build a mapping DataFrame
mapping_df = get_gene_mapping(gene_annotation, prob_col=probe_col, gene_col=gene_col)

# 3. Convert probe-level data to gene-level data
#    Using the debugged function that preserves numeric Entrez IDs
gene_data = apply_gene_mapping_entrez(gene_data, mapping_df)

# Check resulting shape and index
print("Mapped gene_data shape:", gene_data.shape)
print("First 10 gene identifiers in mapped data:", gene_data.index[:10].tolist())
# STEP7
# 1. Normalize the obtained gene data with the 'normalize_gene_symbols_in_index' function from the library.
normalized_gene_data = normalize_gene_symbols_in_index(gene_data)
normalized_gene_data.to_csv(out_gene_data_file)

# 2. Link the clinical and genetic data with the 'geo_link_clinical_genetic_data' function from the library.
#    Replace 'df_clinical' with the correct clinical DataFrame variable 'selected_clinical_df'.
linked_data = geo_link_clinical_genetic_data(selected_clinical_df, normalized_gene_data)

# 3. Handle missing values in the linked data
linked_data = handle_missing_values(linked_data, trait)

# 4. Determine whether the trait and some demographic features are severely biased, and remove biased features.
is_trait_biased, unbiased_linked_data = judge_and_remove_biased_features(linked_data, trait)

# 5. Conduct quality check and save the cohort information, passing the final unbiased data.
is_usable = validate_and_save_cohort_info(
    is_final=True,
    cohort=cohort,
    info_path=json_path,
    is_gene_available=True,
    is_trait_available=True,
    is_biased=is_trait_biased,
    df=unbiased_linked_data
)

# 6. If the linked data is usable, save it as a CSV file to 'out_data_file'.
if is_usable:
    unbiased_linked_data.to_csv(out_data_file)