output/preprocess/Allergies/code/GSE203196.py

# Path Configuration
from tools.preprocess import *

# Processing context
trait = "Allergies"
cohort = "GSE203196"

# Input paths
in_trait_dir = "../DATA/GEO/Allergies"
in_cohort_dir = "../DATA/GEO/Allergies/GSE203196"

# Output paths
out_data_file = "./output/z1/preprocess/Allergies/GSE203196.csv"
out_gene_data_file = "./output/z1/preprocess/Allergies/gene_data/GSE203196.csv"
out_clinical_data_file = "./output/z1/preprocess/Allergies/clinical_data/GSE203196.csv"
json_path = "./output/z1/preprocess/Allergies/cohort_info.json"


# Step 1: Initial Data Loading
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)

# Step 2: Dataset Analysis and Clinical Feature Extraction
import os

# Step 1: Determine gene availability
is_gene_available = True  # Affymetrix transcriptomic studies imply gene expression microarray data is available.

# Step 2: Determine variable availability based on provided sample characteristics dictionary
trait_row = 4    # 'allergy: severe/mild/control'
age_row = 3      # 'age: <number>'
gender_row = 1   # 'gender: F/M'

# Step 2.2: Define conversion functions
def _extract_value(x):
    if x is None:
        return None
    s = str(x).strip()
    if ":" in s:
        s = s.split(":", 1)[1].strip()
    return s if s != "" else None

def convert_trait(x):
    v = _extract_value(x)
    if v is None:
        return None
    v_low = v.lower()
    # Binary allergic status: control -> 0; mild/severe -> 1
    if v_low in {"control", "ctrl", "healthy", "non-allergy", "non allergy", "nonallergy"}:
        return 0
    if v_low in {"allergy", "allergic", "mild", "severe"}:
        return 1
    # Heuristic: unknown strings containing 'control' or 'allerg'
    if "control" in v_low:
        return 0
    if "allerg" in v_low:
        return 1
    return None

def convert_age(x):
    v = _extract_value(x)
    if v is None:
        return None
    # Keep only digits and possible decimal point
    import re
    m = re.search(r"[-+]?\d+(\.\d+)?", v)
    if not m:
        return None
    try:
        return float(m.group())
    except Exception:
        return None

def convert_gender(x):
    v = _extract_value(x)
    if v is None:
        return None
    v_low = v.lower()
    if v_low in {"f", "female", "woman", "women"}:
        return 0
    if v_low in {"m", "male", "man", "men"}:
        return 1
    return None

# Step 3: Initial validation and save metadata
is_trait_available = trait_row is not None
_ = 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
)

# Step 4: Clinical feature extraction (only if trait data is available)
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 = preview_df(selected_clinical_df)
    print(preview)
    # Save clinical features
    os.makedirs(os.path.dirname(out_clinical_data_file), exist_ok=True)
    selected_clinical_df.to_csv(out_clinical_data_file, index=True)

# Step 3: Gene Data Extraction
# 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])

# Step 4: Gene Identifier Review
import os
import re
import pandas as pd

def infer_requires_mapping_from_ids(ids):
    if not ids:
        return True
    n = len(ids)
    ids = [str(x) for x in ids]
    numeric_only = sum(s.isdigit() for s in ids) / n
    has_vendor_prefix = sum(bool(re.match(r'^(ILMN_|A_|AFFX|ENS[A-Z]*|NM_|NR_|XM_|XR_)', s)) for s in ids) / n
    many_underscores = sum('_' in s for s in ids) / n
    # Heuristic: if majority are numeric-only or vendor/platform-style, mapping is required
    if (numeric_only > 0.5) or (has_vendor_prefix > 0.3) or (many_underscores > 0.5):
        return True
    # Otherwise, check if they resemble HGNC symbols (alphanumeric, mostly uppercase, few special chars)
    def looks_like_symbol(s):
        if s.isdigit():
            return False
        if len(s) > 25:
            return False
        # Allowed chars: letters, digits, hyphen, dot
        if not re.match(r'^[A-Za-z0-9\.\-]+$', s):
            return False
        # Must contain at least one letter
        if not re.search(r'[A-Za-z]', s):
            return False
        return True
    symbol_like = sum(looks_like_symbol(s) for s in ids) / n
    return symbol_like < 0.5

gene_ids_sample = ['16657436', '16657440', '16657445', '16657447', '16657450',
                   '16657469', '16657473', '16657476', '16657480', '16657485',
                   '16657489', '16657492', '16657502', '16657506', '16657509',
                   '16657514', '16657527', '16657529', '16657534', '16657554']

ids_to_check = None
if os.path.exists(out_gene_data_file):
    try:
        df_gene = pd.read_csv(out_gene_data_file, index_col=0)
        ids_to_check = df_gene.index.astype(str).tolist()
    except Exception:
        ids_to_check = gene_ids_sample
else:
    ids_to_check = gene_ids_sample

requires_gene_mapping = infer_requires_mapping_from_ids(ids_to_check)
print(f"requires_gene_mapping = {str(requires_gene_mapping)}")

# Step 5: Gene Annotation
# 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
# Ensure required dataframes exist from previous steps
try:
    gene_annotation
except NameError:
    gene_annotation = get_gene_annotation(soft_file)

try:
    probe_data = gene_data
except NameError:
    probe_data = get_genetic_data(matrix_file)

# 1-2. Decide columns: probe IDs in annotation match expression IDs ('ID'); gene symbols info in 'gene_assignment'
probe_col = 'ID'
gene_col = 'gene_assignment'
mapping_df = get_gene_mapping(gene_annotation, prob_col=probe_col, gene_col=gene_col)

# 3. Apply mapping to convert probe-level to gene-level expression; result named gene_data
gene_data = apply_gene_mapping(probe_data, mapping_df)

# Step 7: Data Normalization and Linking
import os
import pandas as pd

# Ensure clinical features are available in current session
if 'selected_clinical_df' not in locals():
    if os.path.exists(out_clinical_data_file):
        selected_clinical_df = pd.read_csv(out_clinical_data_file, index_col=0)
    else:
        # Fallback: re-extract if necessary (should not happen under normal pipeline)
        selected_clinical_df = geo_select_clinical_features(
            clinical_df=clinical_data,
            trait=trait,
            trait_row=4,
            convert_trait=convert_trait,
            age_row=3,
            convert_age=convert_age,
            gender_row=1,
            convert_gender=convert_gender
        )

# Ensure gene_data (gene-level from mapping) is available
if 'gene_data' not in locals():
    # Recompute from raw files if needed
    try:
        gene_annotation
    except NameError:
        gene_annotation = get_gene_annotation(soft_file)
    try:
        mapping_df
    except NameError:
        mapping_df = get_gene_mapping(gene_annotation, prob_col='ID', gene_col='gene_assignment')
    try:
        probe_data
    except NameError:
        probe_data = get_genetic_data(matrix_file)
    gene_data = apply_gene_mapping(probe_data, mapping_df)

# 1. Normalize gene symbols and save
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
linked_data = geo_link_clinical_genetic_data(selected_clinical_df, normalized_gene_data)

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

# 4. Bias assessment and removal of biased demographics
is_trait_biased, unbiased_linked_data = judge_and_remove_biased_features(linked_data, trait)

# Derive availability flags based on actual data
is_gene_available = normalized_gene_data.shape[0] > 0 and normalized_gene_data.shape[1] > 0
is_trait_available = (trait in selected_clinical_df.index) and (not selected_clinical_df.loc[trait].isna().all())

# Optional note: dataset contains multiple cell types which may be a confounder if not modeled
note = "INFO: Samples span multiple cell types (CD14+, CD3+, platelets); consider including cell type as a covariate in downstream analyses."

# 5. Final validation and metadata saving
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=is_trait_available,
    is_biased=is_trait_biased,
    df=unbiased_linked_data,
    note=note
)

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