Add files using upload-large-folder tool

output/preprocess/Testicular_Cancer/code/GSE42647.py

@@ -0,0 +1,130 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Testicular_Cancer"
+cohort = "GSE42647"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Testicular_Cancer"
+in_cohort_dir = "../DATA/GEO/Testicular_Cancer/GSE42647"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Testicular_Cancer/GSE42647.csv"
+out_gene_data_file = "./output/z6/preprocess/Testicular_Cancer/gene_data/GSE42647.csv"
+out_clinical_data_file = "./output/z6/preprocess/Testicular_Cancer/clinical_data/GSE42647.csv"
+json_path = "./output/z6/preprocess/Testicular_Cancer/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 re
+
+# 1) Gene expression data availability (SuperSeries likely includes expression data; not miRNA-only)
+is_gene_available = True
+
+# 2) Variable availability from Sample Characteristics Dictionary:
+# Given the dictionary shows constant cell line info and cell type for all samples (cell line dataset),
+# there is no varying human trait/age/gender information useful for association analysis.
+trait_row = None
+age_row = None
+gender_row = None
+
+# 2.2) Converters
+
+def _after_colon(x: str) -> str:
+    if x is None:
+        return None
+    if not isinstance(x, str):
+        x = str(x)
+    parts = x.split(":", 1)
+    val = parts[1] if len(parts) > 1 else parts[0]
+    return val.strip().strip('"').strip()
+
+def convert_trait(x):
+    v = _after_colon(x)
+    if v is None or v == "" or v.lower() in {"na", "n/a", "null", "none", "unknown"}:
+        return None
+    vl = v.lower()
+    # Heuristics: presence of cancer terms => 1; explicit normal/control => 0
+    cancer_terms = ["cancer", "carcinoma", "tumor", "tumour", "neoplasm", "malignant", "nt2/d1", "embryonal"]
+    normal_terms = ["normal", "control", "healthy", "non-cancer", "noncancer", "adjacent normal"]
+    if any(t in vl for t in cancer_terms):
+        return 1
+    if any(t in vl for t in normal_terms):
+        return 0
+    return None
+
+def convert_age(x):
+    v = _after_colon(x)
+    if v is None or v == "" or v.lower() in {"na", "n/a", "null", "none", "unknown"}:
+        return None
+    # Extract a plausible age number
+    nums = re.findall(r"\d+\.?\d*", v)
+    if not nums:
+        return None
+    try:
+        age = float(nums[0])
+    except:
+        return None
+    # Basic plausibility for human age
+    if 0 <= age <= 120:
+        return age
+    return None
+
+def convert_gender(x):
+    v = _after_colon(x)
+    if v is None or v == "" or v.lower() in {"na", "n/a", "null", "none", "unknown"}:
+        return None
+    vl = v.lower()
+    # map female -> 0, male -> 1
+    if any(tok in vl for tok in ["female", "f", "woman", "women", "girl"]):
+        return 0
+    if any(tok in vl for tok in ["male", "m", "man", "men", "boy"]):
+        return 1
+    return None
+
+# 3) Save metadata (initial filtering)
+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
+)
+
+# 4) Clinical Feature Extraction (skip because trait_row is None)
+# If trait_row were available:
+if trait_row is not None:
+    selected = 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_df(selected)
+    os.makedirs(os.path.dirname(out_clinical_data_file), exist_ok=True)
+    selected.to_csv(out_clinical_data_file, index=True)

output/preprocess/Testicular_Cancer/code/GSE62523.py

@@ -0,0 +1,197 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Testicular_Cancer"
+cohort = "GSE62523"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Testicular_Cancer"
+in_cohort_dir = "../DATA/GEO/Testicular_Cancer/GSE62523"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Testicular_Cancer/GSE62523.csv"
+out_gene_data_file = "./output/z6/preprocess/Testicular_Cancer/gene_data/GSE62523.csv"
+out_clinical_data_file = "./output/z6/preprocess/Testicular_Cancer/clinical_data/GSE62523.csv"
+json_path = "./output/z6/preprocess/Testicular_Cancer/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
+# 1) Gene Expression Data Availability
+is_gene_available = True  # 18k cDNA microarray gene expression in HMEC-1 cells
+
+# 2) Variable Availability and Data Type Conversion
+# No human subject-level clinical data (cell line experiment), hence not available
+trait_row = None
+age_row = None
+gender_row = None
+
+# Define converters per spec (not used since corresponding rows are None)
+def convert_trait(x):
+    # Trait is Testicular_Cancer; not applicable in HMEC-1 cell line experiment
+    return None
+
+def convert_age(x):
+    # No age data in this series
+    return None
+
+def convert_gender(x):
+    # No gender data in this series
+    return None
+
+# 3) Save Metadata (initial filtering)
+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
+)
+
+# 4) Clinical Feature Extraction
+# Skipped because trait_row is None (no clinical trait data available)
+
+# 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
+# The observed identifiers (e.g., '1.1.1.1') are Enzyme Commission (EC) numbers, not human gene symbols.
+print("requires_gene_mapping = True")
+
+# 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
+# Identify the probe ID column and the gene symbol column in the annotation
+probe_col = 'ID'
+gene_symbol_col = 'Gene symbol'
+
+# Build mapping dataframe from probe IDs to gene symbols
+mapping_df = get_gene_mapping(gene_annotation, prob_col=probe_col, gene_col=gene_symbol_col)
+
+# Apply mapping to convert probe-level data to gene-level expression
+gene_data = apply_gene_mapping(expression_df=gene_data, mapping_df=mapping_df)
+
+# Step 7: Data Normalization and Linking
+# Ensure probe-mapped gene_data exists; if not, rebuild from available artifacts
+if 'gene_data' not in locals() or gene_data is None or getattr(gene_data, 'empty', True):
+    if 'matrix_file' in locals():
+        expr_df = get_genetic_data(matrix_file)
+    else:
+        raise RuntimeError("matrix_file not found; cannot rebuild expression data.")
+    # Ensure mapping is available
+    if 'mapping_df' not in locals() or mapping_df is None or getattr(mapping_df, 'empty', True):
+        if 'gene_annotation' not in locals() or gene_annotation is None or getattr(gene_annotation, 'empty', True):
+            if 'soft_file' in locals():
+                gene_annotation = get_gene_annotation(soft_file)
+            else:
+                raise RuntimeError("soft_file not found; cannot rebuild gene annotation/mapping.")
+        mapping_df = get_gene_mapping(gene_annotation, prob_col='ID', gene_col='Gene symbol')
+    gene_data = apply_gene_mapping(expression_df=expr_df, mapping_df=mapping_df)
+
+# 1) Normalize gene symbols and save gene expression data
+normalized_gene_data = normalize_gene_symbols_in_index(gene_data)
+normalized_gene_data.to_csv(out_gene_data_file)
+
+# Determine if clinical/trait data are available
+is_trait_available = False
+if 'trait_row' in locals() and trait_row is not None:
+    is_trait_available = True
+elif 'selected_clinical_data' in locals():
+    is_trait_available = True
+
+# 2-6) Conditional pipeline: only if trait/clinical data available
+if is_trait_available:
+    # If selected_clinical_data is not prepared yet but rows/functions exist, try to build it
+    if 'selected_clinical_data' not in locals():
+        if 'clinical_data' in locals() and 'trait_row' in locals() and trait_row is not None and \
+           'convert_trait' in locals():
+            # Optional covariates if available
+            age_r = age_row if ('age_row' in locals()) else None
+            gen_r = gender_row if ('gender_row' in locals()) else None
+            conv_age = convert_age if ('convert_age' in locals()) else None
+            conv_gen = convert_gender if ('convert_gender' in locals()) else None
+            selected_clinical_data = geo_select_clinical_features(
+                clinical_df=clinical_data,
+                trait=trait,
+                trait_row=trait_row,
+                convert_trait=convert_trait,
+                age_row=age_r,
+                convert_age=conv_age,
+                gender_row=gen_r,
+                convert_gender=conv_gen
+            )
+        else:
+            # Cannot construct clinical data despite is_trait_available being True; fallback to initial filtering
+            _ = validate_and_save_cohort_info(
+                is_final=False,
+                cohort=cohort,
+                info_path=json_path,
+                is_gene_available=True,
+                is_trait_available=False
+            )
+            raise RuntimeError("Trait/clinical data flagged available but cannot be constructed. Recorded metadata and aborting.")
+
+    # 2) Link clinical and genetic data
+    linked_data = geo_link_clinical_genetic_data(selected_clinical_data, normalized_gene_data)
+
+    # 3) Missing value handling
+    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)
+
+    # 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=True,
+        is_trait_available=True,
+        is_biased=is_trait_biased,
+        df=unbiased_linked_data,
+        note="INFO: Linked dataset processed with standard QC."
+    )
+
+    # 6) Save linked data if usable
+    if is_usable:
+        unbiased_linked_data.to_csv(out_data_file)
+else:
+    # No human subject-level trait data (cell line experiment); skip linking and final validation.
+    print("No clinical trait data available (cell-line experiment). Skipping linking/QC; saving gene data only.")
+    _ = validate_and_save_cohort_info(
+        is_final=False,
+        cohort=cohort,
+        info_path=json_path,
+        is_gene_available=True,
+        is_trait_available=False
+    )

output/preprocess/Testicular_Cancer/code/TCGA.py

@@ -0,0 +1,250 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Testicular_Cancer"
+
+# Input paths
+tcga_root_dir = "../DATA/TCGA"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Testicular_Cancer/TCGA.csv"
+out_gene_data_file = "./output/z6/preprocess/Testicular_Cancer/gene_data/TCGA.csv"
+out_clinical_data_file = "./output/z6/preprocess/Testicular_Cancer/clinical_data/TCGA.csv"
+json_path = "./output/z6/preprocess/Testicular_Cancer/cohort_info.json"
+
+
+# Step 1: Initial Data Loading
+import os
+import pandas as pd
+
+# Step 1: Select the most relevant TCGA cohort directory for the trait
+subdirs = [d for d in os.listdir(tcga_root_dir) if os.path.isdir(os.path.join(tcga_root_dir, d))]
+priority_keywords = [
+    'testicular',
+    '(tgct',
+    'tgct',
+    'testis',
+]
+
+selected_dir = None
+lname_map = {d: d.lower() for d in subdirs}
+for kw in priority_keywords:
+    matches = [d for d, ld in lname_map.items() if kw in ld]
+    if matches:
+        # Choose the most specific (longest name) to bias towards explicit matches
+        selected_dir = sorted(matches, key=len, reverse=True)[0]
+        break
+
+if selected_dir is None:
+    # No suitable directory found; record and stop
+    _ = validate_and_save_cohort_info(
+        is_final=False,
+        cohort="TCGA",
+        info_path=json_path,
+        is_gene_available=False,
+        is_trait_available=False
+    )
+    print("No suitable TCGA cohort directory found for the trait. Skipping.")
+    clinical_df = pd.DataFrame()
+    genetic_df = pd.DataFrame()
+else:
+    cohort_dir = os.path.join(tcga_root_dir, selected_dir)
+
+    # Step 2: Identify clinical and genetic file paths
+    clinical_file_path, genetic_file_path = tcga_get_relevant_filepaths(cohort_dir)
+
+    # Step 3: Load both files as DataFrames
+    def read_tsv(path):
+        compression = 'gzip' if path.endswith('.gz') else None
+        return pd.read_csv(path, sep='\t', index_col=0, low_memory=False, compression=compression)
+
+    clinical_df = read_tsv(clinical_file_path)
+    genetic_df = read_tsv(genetic_file_path)
+
+    # Step 4: Print column names of the clinical data
+    print(clinical_df.columns.tolist())
+
+# Step 2: Find Candidate Demographic Features
+import os
+import re
+import pandas as pd
+
+# Given column list from the previous step
+all_columns = ['_INTEGRATION', '_PATIENT', '_cohort', '_primary_disease', '_primary_site', 'age_at_initial_pathologic_diagnosis', 'bcr_followup_barcode', 'bcr_patient_barcode', 'bcr_sample_barcode', 'bilateral_diagnosis_timing_type', 'clinical_M', 'clinical_N', 'clinical_T', 'clinical_stage', 'days_to_bilateral_tumor_dx', 'days_to_birth', 'days_to_collection', 'days_to_death', 'days_to_initial_pathologic_diagnosis', 'days_to_last_followup', 'days_to_new_tumor_event_after_initial_treatment', 'days_to_post_orchi_serum_test', 'days_to_pre_orchi_serum_test', 'eastern_cancer_oncology_group', 'family_history_other_cancer', 'family_history_testicular_cancer', 'family_member_relationship_type', 'first_treatment_success', 'form_completion_date', 'gender', 'histological_percentage', 'histological_type', 'history_fertility', 'history_hypospadias', 'history_of_neoadjuvant_treatment', 'history_of_undescended_testis', 'icd_10', 'icd_o_3_histology', 'icd_o_3_site', 'igcccg_stage', 'informed_consent_verified', 'init_pathology_dx_method_other', 'initial_pathologic_diagnosis_method', 'initial_weight', 'intratubular_germ_cell_neoplasm', 'is_ffpe', 'karnofsky_performance_score', 'laterality', 'level_of_non_descent', 'lost_follow_up', 'lymphovascular_invasion_present', 'molecular_test_result', 'new_neoplasm_event_occurrence_anatomic_site', 'new_neoplasm_event_type', 'new_neoplasm_occurrence_anatomic_site_text', 'new_tumor_event_after_initial_treatment', 'oct_embedded', 'other_dx', 'pathologic_M', 'pathologic_N', 'pathologic_T', 'pathologic_stage', 'pathology_report_file_name', 'patient_death_reason', 'patient_id', 'person_neoplasm_cancer_status', 'post_orchi_afp', 'post_orchi_hcg', 'post_orchi_ldh', 'post_orchi_lh', 'post_orchi_lymph_node_dissection', 'post_orchi_testosterone', 'postoperative_rx_tx', 'postoperative_tx', 'pre_orchi_afp', 'pre_orchi_hcg', 'pre_orchi_ldh', 'pre_orchi_lh', 'pre_orchi_testosterone', 'primary_therapy_outcome_success', 'radiation_therapy', 'relation_testicular_cancer', 'relative_family_cancer_hx_text', 'sample_type', 'sample_type_id', 'serum_markers', 'source_of_patient_death_reason', 'synchronous_tumor_histology_pct', 'synchronous_tumor_histology_type', 'system_version', 'testis_tumor_macroextent', 'testis_tumor_macroextent_other', 'testis_tumor_microextent', 'tissue_prospective_collection_indicator', 'tissue_retrospective_collection_indicator', 'tissue_source_site', 'tumor_tissue_site', 'undescended_testis_corrected', 'undescended_testis_corrected_age', 'undescended_testis_method_left', 'undescended_testis_method_right', 'vial_number', 'vital_status', 'year_of_initial_pathologic_diagnosis', '_GENOMIC_ID_TCGA_TGCT_PDMRNAseqCNV', '_GENOMIC_ID_TCGA_TGCT_exp_HiSeqV2_exon', '_GENOMIC_ID_TCGA_TGCT_mutation_ucsc_maf_gene', '_GENOMIC_ID_TCGA_TGCT_exp_HiSeqV2', '_GENOMIC_ID_TCGA_TGCT_exp_HiSeqV2_PANCAN', '_GENOMIC_ID_TCGA_TGCT_hMethyl450', '_GENOMIC_ID_TCGA_TGCT_gistic2', '_GENOMIC_ID_data/public/TCGA/TGCT/miRNA_HiSeq_gene', '_GENOMIC_ID_TCGA_TGCT_gistic2thd', '_GENOMIC_ID_TCGA_TGCT_mutation_bcm_gene', '_GENOMIC_ID_TCGA_TGCT_miRNA_HiSeq', '_GENOMIC_ID_TCGA_TGCT_mutation_broad_gene', '_GENOMIC_ID_TCGA_TGCT_PDMRNAseq', '_GENOMIC_ID_TCGA_TGCT_RPPA', '_GENOMIC_ID_TCGA_TGCT_exp_HiSeqV2_percentile', '_GENOMIC_ID_TCGA_TGCT_mutation_bcgsc_gene']
+
+# Identify candidate age columns
+def is_age_col(col: str) -> bool:
+    c = col.lower()
+    if re.search(r'(^|_)age($|_)', c):
+        return True
+    if 'age_at' in c:
+        return True
+    # Common proxy for age in TCGA clinical: days_to_birth
+    if c == 'days_to_birth':
+        return True
+    return False
+
+candidate_age_cols = [c for c in all_columns if is_age_col(c)]
+
+# Identify candidate gender columns
+def is_gender_col(col: str) -> bool:
+    c = col.lower()
+    if re.search(r'(^|_)gender($|_)', c):
+        return True
+    if re.search(r'(^|_)sex($|_)', c):
+        return True
+    return False
+
+candidate_gender_cols = [c for c in all_columns if is_gender_col(c)]
+
+print(f"candidate_age_cols = {candidate_age_cols}")
+print(f"candidate_gender_cols = {candidate_gender_cols}")
+
+# Try to load clinical data and preview candidate columns
+clinical_df = None
+try:
+    # Find cohort directory for TGCT within tcga_root_dir
+    cohort_dir = None
+    for d in os.listdir(tcga_root_dir):
+        full = os.path.join(tcga_root_dir, d)
+        if os.path.isdir(full):
+            name = d.lower()
+            if name == 'tgct' or 'tgct' in name or 'testicular' in name:
+                cohort_dir = full
+                break
+    if cohort_dir:
+        clinical_fp, _ = tcga_get_relevant_filepaths(cohort_dir)
+        clinical_df = pd.read_csv(clinical_fp, sep='\t', index_col=0, dtype=str)
+except Exception as e:
+    clinical_df = None
+
+if clinical_df is not None:
+    age_cols_present = [c for c in candidate_age_cols if c in clinical_df.columns]
+    gender_cols_present = [c for c in candidate_gender_cols if c in clinical_df.columns]
+    if age_cols_present:
+        age_preview = preview_df(clinical_df[age_cols_present])
+        print(age_preview)
+    if gender_cols_present:
+        gender_preview = preview_df(clinical_df[gender_cols_present])
+        print(gender_preview)
+
+# Step 3: Select Demographic Features
+# Select age and gender columns from candidate lists based on data quality and interpretability
+try:
+    age_candidates = candidate_age_cols
+except NameError:
+    age_candidates = []
+
+try:
+    gender_candidates = candidate_gender_cols
+except NameError:
+    gender_candidates = []
+
+# Preference order for age: direct age in years > derived (days_to_birth) > others
+age_col = None
+if 'age_at_initial_pathologic_diagnosis' in age_candidates:
+    age_col = 'age_at_initial_pathologic_diagnosis'
+elif 'days_to_birth' in age_candidates:
+    age_col = 'days_to_birth'
+elif len(age_candidates) > 0:
+    age_col = age_candidates[0]
+
+# Preference for gender: 'gender' if available
+gender_col = 'gender' if 'gender' in gender_candidates else (gender_candidates[0] if len(gender_candidates) > 0 else None)
+
+# Print selected columns
+print(f"Selected age_col: {age_col}")
+print(f"Selected gender_col: {gender_col}")
+
+# Attempt to print first-5 preview values if preview dictionaries exist
+def _print_preview(col_name: str, preview_dict_varnames):
+    values = None
+    for varname in preview_dict_varnames:
+        if varname in globals():
+            v = globals()[varname]
+            if isinstance(v, dict) and col_name in v:
+                values = v[col_name]
+                break
+    if values is not None:
+        print(f"{col_name} sample values (first 5): {values}")
+
+# Common potential variable names used in previous steps for previews
+age_preview_varnames = [
+    'age_preview_dict', 'age_candidate_values', 'age_candidates_preview',
+    'age_values_dict', 'age_candidates_dict'
+]
+gender_preview_varnames = [
+    'gender_preview_dict', 'gender_candidate_values', 'gender_candidates_preview',
+    'gender_values_dict', 'gender_candidates_dict'
+]
+
+if age_col is not None:
+    _print_preview(age_col, age_preview_varnames)
+if gender_col is not None:
+    _print_preview(gender_col, gender_preview_varnames)
+
+# Step 4: Feature Engineering and Validation
+import os
+import re
+import pandas as pd
+
+# 1) Extract and standardize clinical features (trait, Age, Gender)
+age_in_cols = (age_col is not None) and (age_col in clinical_df.columns)
+gender_in_cols = (gender_col is not None) and (gender_col in clinical_df.columns)
+selected_clinical_df = tcga_select_clinical_features(
+    clinical_df,
+    trait=trait,
+    age_col=age_col if age_in_cols else None,
+    gender_col=gender_col if gender_in_cols else None
+)
+
+# 2) Normalize gene symbols in gene expression data and save
+gene_df = genetic_df.copy()
+
+# Auto-detect orientation so that gene_df index are gene symbols and columns are samples
+barcode_re = re.compile(r'^TCGA-[A-Z0-9]{2}-[A-Z0-9]{4}')
+sample_n_idx = min(len(gene_df.index), 100)
+sample_n_cols = min(len(gene_df.columns), 100)
+idx_barcodes = sum(1 for v in list(gene_df.index)[:sample_n_idx] if barcode_re.match(str(v)))
+col_barcodes = sum(1 for v in list(gene_df.columns)[:sample_n_cols] if barcode_re.match(str(v)))
+# If index looks like barcodes, transpose to make genes as index
+if idx_barcodes > col_barcodes:
+    gene_df = gene_df.T
+
+# Ensure numeric matrix
+gene_df = gene_df.apply(pd.to_numeric, errors='coerce')
+
+# Normalize gene symbols (removes unrecognized, aggregates synonyms)
+gene_df_norm = normalize_gene_symbols_in_index(gene_df)
+
+# Save normalized gene expression matrix
+os.makedirs(os.path.dirname(out_gene_data_file), exist_ok=True)
+gene_df_norm.to_csv(out_gene_data_file)
+
+# 3) Link clinical and genetic data on sample IDs
+linked_data = pd.concat([selected_clinical_df, gene_df_norm.T], axis=1, join='inner')
+
+# 4) Handle missing values
+processed_df = handle_missing_values(linked_data, trait_col=trait)
+
+# 5) Determine bias and remove biased demographic features
+is_biased, processed_df = judge_and_remove_biased_features(processed_df, trait=trait)
+
+# 6) Final validation and save cohort info
+note = "INFO: Gene symbols normalized via NCBI synonyms; gene matrix orientation auto-detected; samples linked by intersection."
+is_usable = validate_and_save_cohort_info(
+    is_final=True,
+    cohort="TCGA",
+    info_path=json_path,
+    is_gene_available=True,
+    is_trait_available=True,
+    is_biased=is_biased,
+    df=processed_df,
+    note=note
+)
+
+# 7) Save linked data if usable
+if is_usable:
+    os.makedirs(os.path.dirname(out_data_file), exist_ok=True)
+    processed_df.to_csv(out_data_file)

output/preprocess/Testicular_Cancer/cohort_info.json

@@ -1,32 +1 @@
-{
-  "GSE62523": {
-    "is_usable": false,
-    "is_gene_available": true,
-    "is_trait_available": false,
-    "is_available": false,
-    "is_biased": null,
-    "has_age": null,
-    "has_gender": null,
-    "sample_size": null
-  },
-  "GSE42647": {
-    "is_usable": false,
-    "is_gene_available": true,
-    "is_trait_available": true,
-    "is_available": true,
-    "is_biased": true,
-    "has_age": false,
-    "has_gender": false,
-    "sample_size": 12
-  },
-  "TCGA": {
-    "is_usable": false,
-    "is_gene_available": true,
-    "is_trait_available": true,
-    "is_available": true,
-    "is_biased": true,
-    "has_age": true,
-    "has_gender": false,
-    "sample_size": 156
-  }
-}
+{"GSE62523": {"is_usable": false, "is_gene_available": true, "is_trait_available": false, "is_available": false, "is_biased": null, "has_age": null, "has_gender": null, "sample_size": null, "note": null}, "GSE42647": {"is_usable": false, "is_gene_available": true, "is_trait_available": false, "is_available": false, "is_biased": null, "has_age": null, "has_gender": null, "sample_size": null, "note": null}, "TCGA": {"is_usable": false, "is_gene_available": true, "is_trait_available": true, "is_available": true, "is_biased": true, "has_age": true, "has_gender": false, "sample_size": 156, "note": "INFO: Gene symbols normalized via NCBI synonyms; gene matrix orientation auto-detected; samples linked by intersection."}}

output/preprocess/Thymoma/clinical_data/GSE131027.csv

@@ -1,2 +1,2 @@
 ,GSM3759992,GSM3759993,GSM3759994,GSM3759995,GSM3759996,GSM3759997,GSM3759998,GSM3759999,GSM3760000,GSM3760001,GSM3760002,GSM3760003,GSM3760004,GSM3760005,GSM3760006,GSM3760007,GSM3760008,GSM3760009,GSM3760010,GSM3760011,GSM3760012,GSM3760013,GSM3760014,GSM3760015,GSM3760016,GSM3760017,GSM3760018,GSM3760019,GSM3760020,GSM3760021,GSM3760022,GSM3760023,GSM3760024,GSM3760025,GSM3760026,GSM3760027,GSM3760028,GSM3760029,GSM3760030,GSM3760031,GSM3760032,GSM3760033,GSM3760034,GSM3760035,GSM3760036,GSM3760037,GSM3760038,GSM3760039,GSM3760040,GSM3760041,GSM3760042,GSM3760043,GSM3760044,GSM3760045,GSM3760046,GSM3760047,GSM3760048,GSM3760049,GSM3760050,GSM3760051,GSM3760052,GSM3760053,GSM3760054,GSM3760055,GSM3760056,GSM3760057,GSM3760058,GSM3760059,GSM3760060,GSM3760061,GSM3760062,GSM3760063,GSM3760064,GSM3760065,GSM3760066,GSM3760067,GSM3760068,GSM3760069,GSM3760070,GSM3760071,GSM3760072,GSM3760073,GSM3760074,GSM3760075,GSM3760076,GSM3760077,GSM3760078,GSM3760079,GSM3760080,GSM3760081,GSM3760082,GSM3760083
-Thymoma,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0
+Thymoma,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,1.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0

output/preprocess/Thymoma/clinical_data/TCGA.csv

@@ -0,0 +1,127 @@
+sampleID,Thymoma,Age,Gender
+TCGA-3G-AB0O-01,1,47.98904859685147,1
+TCGA-3G-AB0Q-01,1,66.9678302532512,1
+TCGA-3G-AB0T-01,1,45.79876796714579,1
+TCGA-3G-AB14-01,1,51.28815879534565,1
+TCGA-3G-AB19-01,1,76.18617385352498,0
+TCGA-3Q-A9WF-01,1,71.50444900752909,1
+TCGA-3S-A8YW-01,1,63.115674195756334,1
+TCGA-3S-AAYX-01,1,55.715263518138265,0
+TCGA-3T-AA9L-01,1,31.786447638603697,0
+TCGA-4V-A9QI-01,1,68.94182067077344,1
+TCGA-4V-A9QJ-01,1,64.3668720054757,1
+TCGA-4V-A9QL-01,1,58.08624229979466,0
+TCGA-4V-A9QM-01,1,39.90691307323751,1
+TCGA-4V-A9QN-01,1,40.5886379192334,0
+TCGA-4V-A9QQ-01,1,65.41820670773443,0
+TCGA-4V-A9QR-01,1,51.26078028747433,0
+TCGA-4V-A9QS-01,1,53.23203285420944,0
+TCGA-4V-A9QT-01,1,79.26351813826146,0
+TCGA-4V-A9QU-01,1,76.3750855578371,0
+TCGA-4V-A9QW-01,1,62.3709787816564,1
+TCGA-4V-A9QX-01,1,77.34428473648187,0
+TCGA-4X-A9F9-01,1,57.95482546201232,0
+TCGA-4X-A9FA-01,1,52.106776180698155,1
+TCGA-4X-A9FB-01,1,44.284736481861735,1
+TCGA-4X-A9FC-01,1,50.42573579739904,0
+TCGA-4X-A9FD-01,1,43.71800136892539,0
+TCGA-5G-A9ZZ-01,1,52.33949349760438,0
+TCGA-5K-AAAP-01,1,54.59000684462697,1
+TCGA-5U-AB0D-01,1,71.6413415468857,0
+TCGA-5U-AB0E-01,1,62.19301848049281,1
+TCGA-5U-AB0F-01,1,58.893908281998634,0
+TCGA-5V-A9RR-01,1,67.53730321697468,1
+TCGA-X7-A8D6-01,1,48.65434633812457,0
+TCGA-X7-A8D6-11,0,48.65434633812457,0
+TCGA-X7-A8D7-01,1,46.98425735797399,0
+TCGA-X7-A8D7-11,0,46.98425735797399,0
+TCGA-X7-A8D8-01,1,53.22655715263518,1
+TCGA-X7-A8D9-01,1,,0
+TCGA-X7-A8DB-01,1,71.27994524298425,0
+TCGA-X7-A8DC-01,1,68.62696783025325,1
+TCGA-X7-A8DD-01,1,54.677618069815196,0
+TCGA-X7-A8DE-01,1,35.474332648870636,0
+TCGA-X7-A8DF-01,1,40.53114305270363,1
+TCGA-X7-A8DG-01,1,47.95893223819302,1
+TCGA-X7-A8DI-01,1,57.73032169746749,0
+TCGA-X7-A8DJ-01,1,68.56125941136209,1
+TCGA-X7-A8M0-01,1,61.86447638603696,0
+TCGA-X7-A8M1-01,1,51.54277891854894,1
+TCGA-X7-A8M3-01,1,64.93634496919918,0
+TCGA-X7-A8M4-01,1,41.25941136208077,1
+TCGA-X7-A8M5-01,1,65.80698151950719,0
+TCGA-X7-A8M6-01,1,43.11841204654346,1
+TCGA-X7-A8M7-01,1,26.872005475701574,1
+TCGA-X7-A8M8-01,1,45.83436002737851,1
+TCGA-XH-A853-01,1,67.23066392881589,1
+TCGA-XM-A8R8-01,1,67.09924709103353,0
+TCGA-XM-A8R9-01,1,71.66598220396989,0
+TCGA-XM-A8RB-01,1,68.66529774127311,0
+TCGA-XM-A8RC-01,1,67.51540041067761,1
+TCGA-XM-A8RD-01,1,69.54688569472964,0
+TCGA-XM-A8RE-01,1,54.49144421629021,0
+TCGA-XM-A8RF-01,1,84.58042436687201,0
+TCGA-XM-A8RG-01,1,67.90417522245038,1
+TCGA-XM-A8RH-01,1,60.04380561259411,0
+TCGA-XM-A8RI-01,1,77.33333333333333,1
+TCGA-XM-A8RL-01,1,44.85694729637235,1
+TCGA-XM-AAZ1-01,1,42.116358658453116,1
+TCGA-XM-AAZ2-01,1,73.54140999315537,1
+TCGA-XM-AAZ3-01,1,64.65982203969884,1
+TCGA-XU-A92O-01,1,63.4798083504449,1
+TCGA-XU-A92Q-01,1,50.050650239561946,1
+TCGA-XU-A92R-01,1,76.58042436687201,0
+TCGA-XU-A92T-01,1,81.07597535934292,0
+TCGA-XU-A92U-01,1,67.04175222450377,1
+TCGA-XU-A92V-01,1,72.95277207392198,0
+TCGA-XU-A92W-01,1,61.74948665297741,1
+TCGA-XU-A92X-01,1,17.99315537303217,1
+TCGA-XU-A92Y-01,1,71.24161533196441,0
+TCGA-XU-A92Z-01,1,76.9801505817933,1
+TCGA-XU-A930-01,1,43.19507186858316,1
+TCGA-XU-A931-01,1,61.4154688569473,0
+TCGA-XU-A932-01,1,37.28405201916495,1
+TCGA-XU-A933-01,1,52.77481177275838,0
+TCGA-XU-A936-01,1,65.5003422313484,0
+TCGA-XU-AAXV-01,1,45.399041752224505,0
+TCGA-XU-AAXW-01,1,39.44695414099932,0
+TCGA-XU-AAXX-01,1,44.53114305270363,0
+TCGA-XU-AAXY-01,1,43.13757700205339,1
+TCGA-XU-AAXZ-01,1,63.85763175906913,1
+TCGA-XU-AAY0-01,1,61.998631074606436,0
+TCGA-XU-AAY1-01,1,51.408624229979466,0
+TCGA-YT-A95D-01,1,50.275154004106774,0
+TCGA-YT-A95E-01,1,52.49007529089665,1
+TCGA-YT-A95F-01,1,55.75085557837098,0
+TCGA-YT-A95G-01,1,61.713894592744694,0
+TCGA-YT-A95H-01,1,69.4757015742642,1
+TCGA-ZB-A961-01,1,69.59342915811088,1
+TCGA-ZB-A962-01,1,64.23271731690623,1
+TCGA-ZB-A963-01,1,46.5160848733744,1
+TCGA-ZB-A964-01,1,70.80355920602327,1
+TCGA-ZB-A965-01,1,75.61122518822724,0
+TCGA-ZB-A966-01,1,78.36002737850787,0
+TCGA-ZB-A969-01,1,72.43805612594113,0
+TCGA-ZB-A96A-01,1,53.18001368925393,1
+TCGA-ZB-A96B-01,1,51.723477070499655,1
+TCGA-ZB-A96C-01,1,77.26762491444217,0
+TCGA-ZB-A96D-01,1,52.76933607118412,0
+TCGA-ZB-A96E-01,1,60.48459958932238,1
+TCGA-ZB-A96F-01,1,39.27173169062286,1
+TCGA-ZB-A96G-01,1,59.13210130047912,1
+TCGA-ZB-A96H-01,1,73.05407255304586,1
+TCGA-ZB-A96I-01,1,71.75359342915812,0
+TCGA-ZB-A96K-01,1,69.80971937029432,1
+TCGA-ZB-A96L-01,1,70.94866529774127,0
+TCGA-ZB-A96M-01,1,56.46269678302532,0
+TCGA-ZB-A96O-01,1,54.16016427104723,0
+TCGA-ZB-A96P-01,1,73.56605065023956,1
+TCGA-ZB-A96Q-01,1,49.96303901437371,1
+TCGA-ZB-A96R-01,1,38.26694045174538,1
+TCGA-ZB-A96V-01,1,44.47638603696099,1
+TCGA-ZC-AAA7-01,1,63.696098562628336,1
+TCGA-ZC-AAAA-01,1,39.50444900752909,1
+TCGA-ZC-AAAF-01,1,74.19301848049281,0
+TCGA-ZC-AAAH-01,1,52.41889117043121,1
+TCGA-ZL-A9V6-01,1,61.45653661875428,1
+TCGA-ZT-A8OM-01,1,73.71115674195757,0

output/preprocess/Thymoma/code/GSE131027.py

@@ -0,0 +1,171 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Thymoma"
+cohort = "GSE131027"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Thymoma"
+in_cohort_dir = "../DATA/GEO/Thymoma/GSE131027"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Thymoma/GSE131027.csv"
+out_gene_data_file = "./output/z6/preprocess/Thymoma/gene_data/GSE131027.csv"
+out_clinical_data_file = "./output/z6/preprocess/Thymoma/clinical_data/GSE131027.csv"
+json_path = "./output/z6/preprocess/Thymoma/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
+
+# 1) Gene Expression Data Availability
+is_gene_available = True  # Based on series design indicating expression features (HRD expression, PARP sensitivity)
+
+# 2) Variable Availability and Data Type Conversion
+
+# Trait: Thymoma vs others from 'cancer: ...' -> key 1
+trait_row = 1
+
+# Age/Gender: not available in provided characteristics
+age_row = None
+gender_row = None
+
+def _post_colon_value(x):
+    if x is None:
+        return None
+    s = str(x)
+    # take the substring after the last colon to handle fields with multiple colons
+    if ":" in s:
+        s = s.split(":")[-1]
+    return s.strip()
+
+def convert_trait(x):
+    v = _post_colon_value(x)
+    if not v:
+        return None
+    v_l = v.lower()
+    return 1 if v_l == "thymoma" else 0
+
+# Age and Gender converters are not needed since rows are not available
+convert_age = None
+convert_gender = None
+
+# 3) Save Metadata (initial filtering)
+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
+)
+
+# 4) Clinical Feature Extraction (only if clinical 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
+    )
+    clinical_preview = preview_df(selected_clinical_df, n=5)
+    print(clinical_preview)
+    os.makedirs(os.path.dirname(out_clinical_data_file), exist_ok=True)
+    selected_clinical_df.to_csv(out_clinical_data_file)
+
+# 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
+print("requires_gene_mapping = True")
+
+# 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
+# Identify the appropriate columns in the annotation: 'ID' for probes and 'Gene Symbol' for gene symbols
+mapping_df = get_gene_mapping(gene_annotation, prob_col='ID', gene_col='Gene Symbol')
+
+# Apply mapping to convert probe-level data to gene-level expression
+gene_data = apply_gene_mapping(expression_df=gene_data, mapping_df=mapping_df)
+
+# Step 7: Data Normalization and Linking
+import os
+import pandas as pd
+
+# 1. Normalize the obtained gene data with the 'normalize_gene_symbols_in_index' function from the library.
+os.makedirs(os.path.dirname(out_gene_data_file), exist_ok=True)
+normalized_gene_data = normalize_gene_symbols_in_index(gene_data)
+normalized_gene_data.to_csv(out_gene_data_file)
+
+# Make sure the clinical dataframe is available (load from disk if needed)
+try:
+    selected_clinical_df
+except NameError:
+    selected_clinical_df = pd.read_csv(out_clinical_data_file, index_col=0)
+
+# 2. Link the clinical and genetic data with the 'geo_link_clinical_genetic_data' function from the library.
+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)
+
+# Prepare an informative note about class imbalance using original clinical labels
+trait_series = selected_clinical_df.loc[trait].dropna()
+pos = int((trait_series == 1).sum())
+neg = int((trait_series == 0).sum())
+n_total = pos + neg
+note = f"WARNING: Severe class imbalance for '{trait}': {pos} positive vs {neg} negative (N={n_total})."
+
+# 5. Conduct quality check and save the cohort information.
+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,
+    note=note
+)
+
+# 6. If the linked data is usable, save it as a CSV file to 'out_data_file'.
+if is_usable:
+    os.makedirs(os.path.dirname(out_data_file), exist_ok=True)
+    unbiased_linked_data.to_csv(out_data_file)

output/preprocess/Thymoma/code/GSE29695.py

@@ -0,0 +1,190 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Thymoma"
+cohort = "GSE29695"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Thymoma"
+in_cohort_dir = "../DATA/GEO/Thymoma/GSE29695"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Thymoma/GSE29695.csv"
+out_gene_data_file = "./output/z6/preprocess/Thymoma/gene_data/GSE29695.csv"
+out_clinical_data_file = "./output/z6/preprocess/Thymoma/clinical_data/GSE29695.csv"
+json_path = "./output/z6/preprocess/Thymoma/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
+# 1. Gene Expression Data Availability
+is_gene_available = True  # Whole-genome gene expression on Illumina Human Ref-8 Beadchip per series summary
+
+# 2. Variable Availability and Conversion
+
+# From the provided Sample Characteristics Dictionary, there is no explicit human age or gender.
+# The "trait" for this project is Thymoma, which is constant across human tumor samples in this series.
+# Therefore, treat trait as not available for association purposes (no variation).
+trait_row = None
+age_row = None
+gender_row = None
+
+# Conversion functions (defined for interface completeness; they won't be used since rows are None)
+def _extract_value(cell):
+    if cell is None:
+        return None
+    # Expect "key: value" format; robustly split on the first colon
+    parts = str(cell).split(":", 1)
+    val = parts[1].strip() if len(parts) > 1 else str(cell).strip()
+    # Normalize NA-like tokens
+    if val in {"NA", "na", "Na", "N/A", "n/a", "Unknown", "unknown", ""}:
+        return None
+    return val
+
+def convert_trait(cell):
+    # Not used; placeholder: map presence of thymoma-related descriptors to 1
+    val = _extract_value(cell)
+    if val is None:
+        return None
+    txt = val.lower()
+    if "thymoma" in txt or "thymic" in txt or "tumor" in txt:
+        return 1
+    if "normal" in txt or "control" in txt:
+        return 0
+    return None
+
+def convert_age(cell):
+    val = _extract_value(cell)
+    if val is None:
+        return None
+    # Try to parse numeric age from strings like "45", "45 years", etc.
+    try:
+        # Keep digits and dot
+        import re
+        m = re.search(r"(\d+\.?\d*)", val)
+        return float(m.group(1)) if m else None
+    except Exception:
+        return None
+
+def convert_gender(cell):
+    val = _extract_value(cell)
+    if val is None:
+        return None
+    v = val.strip().lower()
+    # Normalize common gender representations
+    if v in {"female", "f", "woman", "women"}:
+        return 0
+    if v in {"male", "m", "man", "men"}:
+        return 1
+    return None
+
+# 3. Save Metadata (initial filtering)
+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
+)
+
+# 4. Clinical Feature Extraction (skip because trait_row is None)
+# If trait_row were available:
+# selected_clinical = 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)
+# os.makedirs(os.path.dirname(out_clinical_data_file), exist_ok=True)
+# selected_clinical.to_csv(out_clinical_data_file)
+
+# 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
+print("requires_gene_mapping = True")
+
+# 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
+# 1-2. Decide the identifier and gene symbol columns based on the annotation preview
+probe_col = 'ID'       # Matches ILMN_* probe IDs seen in gene expression data
+gene_symbol_col = 'Symbol'  # Standard human gene symbols
+
+# Build the mapping dataframe
+mapping_df = get_gene_mapping(gene_annotation, prob_col=probe_col, gene_col=gene_symbol_col)
+
+# 3. Apply the mapping to convert probe-level data to gene-level expression
+gene_data = apply_gene_mapping(gene_data, mapping_df)
+
+# Step 7: Data Normalization and Linking
+import os
+import pandas as pd
+
+# 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)
+
+# Since trait data is unavailable (trait_row was None in Step 2), skip linking and downstream steps.
+linked_data = None
+is_trait_available = False
+is_gene_available = True
+
+# Prepare a small dummy dataframe to pass final validation without triggering abnormality override
+if normalized_gene_data.shape[1] >= 5:
+    dummy_df = normalized_gene_data.iloc[:1, :5].copy()
+else:
+    dummy_df = pd.DataFrame([[0, 0, 0, 0, 0]], columns=[f"col{i}" for i in range(5)])
+
+is_trait_biased = False  # Placeholder; trait is unavailable
+
+# 5. Final validation and save cohort info
+note = "INFO: Trait data unavailable in this series; skipped linking, missing value handling, and bias checks."
+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=dummy_df,
+    note=note
+)
+
+# 6. Do not save linked data since dataset is not usable (trait unavailable)

output/preprocess/Thymoma/code/GSE42977.py

@@ -0,0 +1,252 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Thymoma"
+cohort = "GSE42977"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Thymoma"
+in_cohort_dir = "../DATA/GEO/Thymoma/GSE42977"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Thymoma/GSE42977.csv"
+out_gene_data_file = "./output/z6/preprocess/Thymoma/gene_data/GSE42977.csv"
+out_clinical_data_file = "./output/z6/preprocess/Thymoma/clinical_data/GSE42977.csv"
+json_path = "./output/z6/preprocess/Thymoma/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
+# 1) Determine gene expression availability
+is_gene_available = True  # Microarray gene-expression dataset per background info
+
+# 2) Determine variable availability
+trait_row = 0  # 'tissue' field contains 'Thymoma' and 'Metastatic Thymoma'
+age_row = None
+gender_row = None
+
+# 2.2) Conversion functions
+def _extract_value(x):
+    if x is None or (isinstance(x, float) and pd.isna(x)):
+        return None
+    s = str(x).strip()
+    if ":" in s:
+        s = s.split(":", 1)[1].strip()
+    if s == "" or s.lower() in {"na", "n/a", "nan", "none", "unknown", "null"}:
+        return None
+    return s
+
+def convert_trait(x):
+    v = _extract_value(x)
+    if v is None:
+        return None
+    v_lower = v.lower()
+    # Positive for any thymoma mention (includes metastatic thymoma)
+    if "thymoma" in v_lower:
+        return 1
+    # All other tissues are considered non-thymoma
+    return 0
+
+def convert_age(x):
+    v = _extract_value(x)
+    if v is None:
+        return None
+    # Extract the first number as age
+    import re
+    m = re.search(r"[-+]?\d*\.?\d+", v)
+    if m:
+        try:
+            return float(m.group())
+        except Exception:
+            return None
+    return None
+
+def convert_gender(x):
+    v = _extract_value(x)
+    if v is None:
+        return None
+    v_lower = v.lower()
+    if v_lower in {"f", "female", "woman", "women"}:
+        return 0
+    if v_lower in {"m", "male", "man", "men"}:
+        return 1
+    return None
+
+# 3) Save metadata (initial filtering)
+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
+)
+
+# 4) Clinical Feature Extraction (only if clinical 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
+    )
+    print(preview_df(selected_clinical_df))
+    os.makedirs(os.path.dirname(out_clinical_data_file), exist_ok=True)
+    selected_clinical_df.to_csv(out_clinical_data_file)
+
+# 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
+
+# Try to get gene identifiers from the saved gene data; fall back to provided sample IDs
+gene_ids = []
+try:
+    if os.path.exists(out_gene_data_file):
+        df_tmp = pd.read_csv(out_gene_data_file, index_col=0)
+        gene_ids = df_tmp.index.astype(str).tolist()
+except Exception:
+    pass
+
+if not gene_ids:
+    gene_ids = [
+        'ILMN_10000', 'ILMN_100000', 'ILMN_100007', 'ILMN_100009', 'ILMN_10001',
+        'ILMN_100010', 'ILMN_10002', 'ILMN_100028', 'ILMN_100030', 'ILMN_100031',
+        'ILMN_100034', 'ILMN_100037', 'ILMN_10004', 'ILMN_10005', 'ILMN_100054',
+        'ILMN_100059', 'ILMN_10006', 'ILMN_100075', 'ILMN_100079', 'ILMN_100083'
+    ]
+
+def needs_mapping(ids):
+    # Common non-gene-symbol identifier patterns
+    patterns = [
+        r'^ILMN_\d+$',             # Illumina probe IDs
+        r'^ENSG\d+(\.\d+)?$',      # Ensembl genes
+        r'^\d+_(at|s_at|x_at|a_at)$',  # Affymetrix probes
+        r'^(NM|NR|XM|XR)_\d+(\.\d+)?$',# RefSeq transcripts
+        r'^cg\d{6,}$',             # CpG probes
+        r'^A_\d+_P\d+$'            # Agilent probes
+    ]
+    comp = [re.compile(p) for p in patterns]
+    matches = 0
+    for x in ids:
+        s = str(x)
+        if any(c.match(s) for c in comp):
+            matches += 1
+    # If majority match known non-symbol patterns, mapping required
+    if len(ids) > 0 and (matches / len(ids)) >= 0.5:
+        return True
+    # Heuristic: underscores are rare in HGNC symbols; many underscores suggest probes/accessions
+    underscore_ratio = sum('_' in str(x) for x in ids) / max(len(ids), 1)
+    if underscore_ratio >= 0.5:
+        return True
+    return False
+
+requires_gene_mapping = needs_mapping(gene_ids)
+print(f"requires_gene_mapping = {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
+# 1-2) Decide mapping columns and build mapping dataframe
+prob_col = 'ID'        # Probe IDs like ILMN_XXXXX
+gene_col = 'Symbol'    # Gene symbols
+
+mapping_df = get_gene_mapping(gene_annotation, prob_col=prob_col, gene_col=gene_col)
+
+# 3) Apply mapping to convert probe-level data to gene-level expression
+gene_data = apply_gene_mapping(expression_df=gene_data, mapping_df=mapping_df)
+
+# Optionally save the mapped gene expression data for later steps
+os.makedirs(os.path.dirname(out_gene_data_file), exist_ok=True)
+gene_data.to_csv(out_gene_data_file)
+
+# Step 7: Data Normalization and Linking
+import os
+import pandas as pd
+
+# 1) Normalize gene symbols and save
+# Ensure we have gene_data from previous steps; if not, reload from disk
+try:
+    gene_data
+except NameError:
+    if os.path.exists(out_gene_data_file):
+        gene_data = pd.read_csv(out_gene_data_file, index_col=0)
+    else:
+        raise RuntimeError("Gene data not found in memory or on disk.")
+
+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
+# Ensure clinical data is available; if not in memory, load from saved file
+try:
+    selected_clinical_df
+except NameError:
+    if os.path.exists(out_clinical_data_file):
+        selected_clinical_df = pd.read_csv(out_clinical_data_file, index_col=0)
+    else:
+        raise RuntimeError("Clinical data not found in memory or on disk.")
+
+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 checks and removal of biased demographics
+is_trait_biased, unbiased_linked_data = judge_and_remove_biased_features(linked_data, trait)
+
+# 5) Final validation and metadata saving
+note = ("INFO: Gene IDs mapped from Illumina ILMN probes to HGNC symbols; "
+        "trait derived from 'tissue' field where any mention of 'Thymoma' is labeled as case (1).")
+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,
+    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)

output/preprocess/Thymoma/code/TCGA.py

@@ -0,0 +1,424 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Thymoma"
+
+# Input paths
+tcga_root_dir = "../DATA/TCGA"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Thymoma/TCGA.csv"
+out_gene_data_file = "./output/z6/preprocess/Thymoma/gene_data/TCGA.csv"
+out_clinical_data_file = "./output/z6/preprocess/Thymoma/clinical_data/TCGA.csv"
+json_path = "./output/z6/preprocess/Thymoma/cohort_info.json"
+
+
+# Step 1: Initial Data Loading
+import os
+import pandas as pd
+
+# Step 1: Select the most relevant TCGA cohort directory for the trait
+all_dirs = [d for d in os.listdir(tcga_root_dir) if os.path.isdir(os.path.join(tcga_root_dir, d))]
+
+def score_dir(name: str, trait_name: str) -> int:
+    nl = name.lower()
+    t = trait_name.lower()
+    score = 0
+    if t in nl:
+        score += 10
+    # Common code/synonym matches
+    if '(thym)' in name:
+        score += 5
+    if 'thym' in nl:
+        score += 2
+    return score
+
+candidates = [(d, score_dir(d, trait)) for d in all_dirs]
+# Keep only positively scored directories
+candidates = [d for d in candidates if d[1] > 0]
+
+selected_dir = None
+if candidates:
+    # Choose the directory with the highest score; in ties, choose the first in sorted order
+    candidates.sort(key=lambda x: (-x[1], x[0]))
+    selected_dir = candidates[0][0]
+
+if selected_dir is None:
+    # No suitable directory found: record and skip
+    validate_and_save_cohort_info(
+        is_final=False,
+        cohort="TCGA",
+        info_path=json_path,
+        is_gene_available=False,
+        is_trait_available=False
+    )
+else:
+    # Step 2: Identify clinical and genetic file paths
+    cohort_dir_path = os.path.join(tcga_root_dir, selected_dir)
+    clinical_path, genetic_path = tcga_get_relevant_filepaths(cohort_dir_path)
+
+    # Step 3: Load both files into DataFrames
+    def read_tsv(path: str) -> pd.DataFrame:
+        comp = 'gzip' if path.endswith('.gz') else None
+        return pd.read_csv(path, sep='\t', index_col=0, low_memory=False, compression=comp)
+
+    clinical_df = read_tsv(clinical_path)
+    genetic_df = read_tsv(genetic_path)
+
+    # Step 4: Print column names of the clinical data
+    print(list(clinical_df.columns))
+
+# Step 2: Find Candidate Demographic Features
+import os
+import re
+import pandas as pd
+
+# Given column list from the previous step
+columns = ['_INTEGRATION', '_PATIENT', '_cohort', '_primary_disease', '_primary_site', 'additional_pharmaceutical_therapy', 'additional_radiation_therapy', 'age_at_initial_pathologic_diagnosis', 'bcr_followup_barcode', 'bcr_patient_barcode', 'bcr_sample_barcode', 'days_to_birth', 'days_to_collection', 'days_to_death', 'days_to_initial_pathologic_diagnosis', 'days_to_last_followup', 'days_to_new_tumor_event_after_initial_treatment', 'form_completion_date', 'gender', 'height', 'histological_type', 'history_myasthenia_gravis', 'history_of_neoadjuvant_treatment', 'icd_10', 'icd_o_3_histology', 'icd_o_3_site', 'informed_consent_verified', 'init_pathology_dx_method_other', 'initial_pathologic_diagnosis_method', 'initial_weight', 'is_ffpe', 'lost_follow_up', 'masaoka_stage', 'new_neoplasm_event_occurrence_anatomic_site', 'new_neoplasm_event_type', 'new_neoplasm_occurrence_anatomic_site_text', 'new_tumor_event_additional_surgery_procedure', 'new_tumor_event_after_initial_treatment', 'oct_embedded', 'other_dx', 'pathology_report_file_name', 'patient_id', 'person_neoplasm_cancer_status', 'post_op_ablation_embolization_tx', 'postoperative_rx_tx', 'radiation_therapy', 'sample_type', 'sample_type_id', 'section_myasthenia_gravis', 'tissue_prospective_collection_indicator', 'tissue_retrospective_collection_indicator', 'tissue_source_site', 'tumor_tissue_site', 'vial_number', 'vital_status', 'weight', 'year_of_initial_pathologic_diagnosis', '_GENOMIC_ID_TCGA_THYM_PDMRNAseqCNV', '_GENOMIC_ID_TCGA_THYM_exp_HiSeqV2_exon', '_GENOMIC_ID_TCGA_THYM_mutation_ucsc_maf_gene', '_GENOMIC_ID_TCGA_THYM_exp_HiSeqV2', '_GENOMIC_ID_TCGA_THYM_exp_HiSeqV2_PANCAN', '_GENOMIC_ID_TCGA_THYM_hMethyl450', '_GENOMIC_ID_TCGA_THYM_gistic2', '_GENOMIC_ID_TCGA_THYM_gistic2thd', '_GENOMIC_ID_data/public/TCGA/THYM/miRNA_HiSeq_gene', '_GENOMIC_ID_TCGA_THYM_mutation_bcm_gene', '_GENOMIC_ID_TCGA_THYM_miRNA_HiSeq', '_GENOMIC_ID_TCGA_THYM_mutation_broad_gene', '_GENOMIC_ID_TCGA_THYM_PDMRNAseq', '_GENOMIC_ID_TCGA_THYM_RPPA', '_GENOMIC_ID_TCGA_THYM_exp_HiSeqV2_percentile', '_GENOMIC_ID_TCGA_THYM_mutation_bcgsc_gene']
+
+# Identify candidate columns with strict matching to avoid false positives like "stage"
+candidate_age_cols = []
+candidate_gender_cols = []
+
+for c in columns:
+    cl = c.lower()
+    is_age = bool(re.search(r'(^|[^a-z])age([^a-z]|$)', cl)) or ('days_to_birth' in cl) or ('birth' in cl)
+    exclude_age = ('stage' in cl) or ('percentage' in cl) or ('percent' in cl) or ('average' in cl)
+    if is_age and not exclude_age:
+        candidate_age_cols.append(c)
+    if ('gender' in cl) or bool(re.search(r'(^|[^a-z])sex([^a-z]|$)', cl)):
+        candidate_gender_cols.append(c)
+
+print(f"candidate_age_cols = {candidate_age_cols}")
+print(f"candidate_gender_cols = {candidate_gender_cols}")
+
+# Load clinical data and preview selected columns if available
+cohort_dir = os.path.join(tcga_root_dir, "THYM")
+if os.path.isdir(cohort_dir):
+    try:
+        clinical_file_path, _ = tcga_get_relevant_filepaths(cohort_dir)
+        clinical_df = pd.read_csv(clinical_file_path, sep='\t', header=0, index_col=0, dtype=str, low_memory=False)
+
+        age_cols_present = [c for c in candidate_age_cols if c in clinical_df.columns]
+        gender_cols_present = [c for c in candidate_gender_cols if c in clinical_df.columns]
+
+        if age_cols_present:
+            print(preview_df(clinical_df[age_cols_present], n=5))
+        if gender_cols_present:
+            print(preview_df(clinical_df[gender_cols_present], n=5))
+    except Exception:
+        pass
+
+# Step 3: Select Demographic Features
+# Initialize defaults
+age_col = None
+gender_col = None
+
+# Helper functions
+def _is_missing(v):
+    if v is None:
+        return True
+    try:
+        import math
+        if isinstance(v, float) and math.isnan(v):
+            return True
+    except Exception:
+        pass
+    s = str(v).strip().lower()
+    return s in {"", "na", "nan", "none", "null"}
+
+def _find_preview_dict(candidates):
+    # Locate a preview dict in globals that contains candidate keys (first-5-values dict from prior step)
+    found = []
+    for name, obj in globals().items():
+        if isinstance(obj, dict) and obj:
+            keys = set(obj.keys())
+            if any(k in keys for k in candidates):
+                found.append(obj)
+    if not found:
+        return None
+    found.sort(key=lambda d: -len(set(d.keys()).intersection(set(candidates))))
+    return found[0]
+
+def _age_plausible_count(vals, colname):
+    # Handle days_to_birth specially by converting to years via abs(days)/365.25
+    col_l = (colname or "").lower()
+    plausible = 0
+    for v in vals:
+        if _is_missing(v):
+            continue
+        if "days_to_birth" in col_l:
+            try:
+                d = float(v)
+            except Exception:
+                # try extracting digits
+                a = tcga_convert_age(v)
+                d = float(a) if a is not None else None
+            if d is None:
+                continue
+            years = abs(d) / 365.25
+            if 0 <= years <= 120:
+                plausible += 1
+        else:
+            a = tcga_convert_age(v)
+            if a is not None and 0 <= a <= 120:
+                plausible += 1
+    return plausible
+
+def _select_from_preview_dict(preview_dict, candidates, is_gender=False):
+    if not isinstance(preview_dict, dict) or not preview_dict:
+        return None
+    usable_cols = [c for c in candidates if c in preview_dict]
+    if not usable_cols:
+        return None
+
+    scored = []
+    for col in usable_cols:
+        vals = preview_dict.get(col, [])
+        non_missing = sum(0 if _is_missing(v) else 1 for v in vals)
+        if is_gender:
+            valid = sum(1 for v in vals if tcga_convert_gender(v) in (0, 1))
+            # Prefer canonical 'gender' on tie
+            preference = 0 if col.lower() == "gender" else 1
+            scored.append((col, valid, non_missing, preference))
+        else:
+            plausible = _age_plausible_count(vals, col)
+            # Prefer 'age_at_initial_pathologic_diagnosis' then 'days_to_birth'
+            pref = 0 if col.lower() == "age_at_initial_pathologic_diagnosis" else (1 if col.lower() == "days_to_birth" else 2)
+            scored.append((col, plausible, non_missing, pref))
+
+    if not scored:
+        return None
+
+    # Sort by: valid/plausible desc, non_missing desc, preference asc
+    scored.sort(key=lambda x: (-x[1], -x[2], x[3]))
+    best_col = scored[0][0]
+    return best_col
+
+def _pick_from_dataframe(candidates, is_gender=False):
+    import pandas as pd
+    dfs = []
+    for name, obj in globals().items():
+        if isinstance(obj, pd.DataFrame) and any(c in obj.columns for c in candidates):
+            dfs.append(obj)
+    if not dfs:
+        return None, None
+    df = sorted(dfs, key=lambda d: -len(d))[0]
+
+    best_col = None
+    best_score = None
+    for col in candidates:
+        if col not in df.columns:
+            continue
+        series = df[col]
+        try:
+            non_missing_ratio = series.notna().mean() if len(series) > 0 else 0.0
+        except Exception:
+            non_missing_ratio = 0.0
+
+        if is_gender:
+            try:
+                valid_ratio = series.apply(lambda v: tcga_convert_gender(v) in (0, 1)).mean() if len(series) > 0 else 0.0
+            except Exception:
+                valid_ratio = 0.0
+            preference = 0 if col.lower() == "gender" else 1
+            score = (valid_ratio, non_missing_ratio, -preference)
+        else:
+            col_l = col.lower()
+            plausible_ratio = 0.0
+            try:
+                if "days_to_birth" in col_l:
+                    # convert to years
+                    s = pd.to_numeric(series, errors="coerce")
+                    years = s.abs() / 365.25
+                    plausible_ratio = years.between(0, 120).mean()
+                else:
+                    ages = series.apply(tcga_convert_age)
+                    plausible_ratio = ages.apply(lambda a: (a is not None) and (0 <= a <= 120)).mean()
+            except Exception:
+                plausible_ratio = 0.0
+            pref_rank = 0 if col_l == "age_at_initial_pathologic_diagnosis" else (1 if col_l == "days_to_birth" else 2)
+            score = (plausible_ratio, non_missing_ratio, - (2 - pref_rank))  # prefer lower pref_rank
+
+        if (best_score is None) or (score > best_score):
+            best_score = score
+            best_col = col
+
+    return best_col, df
+
+# Find the preview dictionaries prepared in the previous step
+age_preview_dict = _find_preview_dict(candidate_age_cols)
+gender_preview_dict = _find_preview_dict(candidate_gender_cols)
+
+# Try selection from preview dicts
+if age_preview_dict:
+    age_col = _select_from_preview_dict(age_preview_dict, candidate_age_cols, is_gender=False)
+
+if gender_preview_dict:
+    gender_col = _select_from_preview_dict(gender_preview_dict, candidate_gender_cols, is_gender=True)
+
+# If still not determined, try using an available DataFrame
+age_df = None
+gender_df = None
+
+if age_col is None and candidate_age_cols:
+    age_col, age_df = _pick_from_dataframe(candidate_age_cols, is_gender=False)
+
+if gender_col is None and candidate_gender_cols:
+    gender_col, gender_df = _pick_from_dataframe(candidate_gender_cols, is_gender=True)
+
+# Fallback to canonical defaults if still None
+if age_col is None:
+    if "age_at_initial_pathologic_diagnosis" in candidate_age_cols:
+        age_col = "age_at_initial_pathologic_diagnosis"
+    elif "days_to_birth" in candidate_age_cols:
+        age_col = "days_to_birth"
+    elif candidate_age_cols:
+        age_col = candidate_age_cols[0]
+    else:
+        age_col = None
+
+if gender_col is None:
+    if "gender" in candidate_gender_cols:
+        gender_col = "gender"
+    elif candidate_gender_cols:
+        gender_col = candidate_gender_cols[0]
+    else:
+        gender_col = None
+
+# Explicitly print out information for chosen columns using the preview dicts if available, otherwise from DataFrame
+print("Chosen age_col:", age_col)
+if age_col is not None:
+    if age_preview_dict and age_col in age_preview_dict:
+        print("Age preview values (first 5):", age_preview_dict[age_col])
+    elif age_df is not None and age_col in getattr(age_df, "columns", []):
+        try:
+            print("Age preview values (first 5):", age_df[age_col].head(5).tolist())
+        except Exception:
+            print("Age preview values (first 5):", None)
+    else:
+        # Try to find any df containing the chosen column for preview
+        try:
+            import pandas as pd
+            for name, obj in globals().items():
+                if isinstance(obj, pd.DataFrame) and age_col in obj.columns:
+                    print("Age preview values (first 5):", obj[age_col].head(5).tolist())
+                    break
+            else:
+                print("Age preview values (first 5):", None)
+        except Exception:
+            print("Age preview values (first 5):", None)
+else:
+    print("Age preview values (first 5):", None)
+
+print("Chosen gender_col:", gender_col)
+if gender_col is not None:
+    if gender_preview_dict and gender_col in gender_preview_dict:
+        print("Gender preview values (first 5):", gender_preview_dict[gender_col])
+    elif gender_df is not None and gender_col in getattr(gender_df, "columns", []):
+        try:
+            print("Gender preview values (first 5):", gender_df[gender_col].head(5).tolist())
+        except Exception:
+            print("Gender preview values (first 5):", None)
+    else:
+        # Try to find any df containing the chosen column for preview
+        try:
+            import pandas as pd
+            for name, obj in globals().items():
+                if isinstance(obj, pd.DataFrame) and gender_col in obj.columns:
+                    print("Gender preview values (first 5):", obj[gender_col].head(5).tolist())
+                    break
+            else:
+                print("Gender preview values (first 5):", None)
+        except Exception:
+            print("Gender preview values (first 5):", None)
+else:
+    print("Gender preview values (first 5):", None)
+
+# Step 4: Feature Engineering and Validation
+import os
+import pandas as pd
+
+# 1) Extract and standardize clinical features
+# Use previously selected columns; if not defined, fall back to None
+try:
+    age_col
+except NameError:
+    age_col = None
+try:
+    gender_col
+except NameError:
+    gender_col = None
+
+selected_clinical = tcga_select_clinical_features(
+    clinical_df=clinical_df,
+    trait=trait,
+    age_col=age_col,
+    gender_col=gender_col
+)
+
+# If age was taken from days_to_birth, convert to years
+if ("Age" in selected_clinical.columns) and (age_col is not None) and (age_col.lower() == "days_to_birth"):
+    s = pd.to_numeric(selected_clinical["Age"], errors="coerce")
+    selected_clinical["Age"] = s.abs() / 365.25
+
+# Save clinical data for transparency (optional but helpful)
+os.makedirs(os.path.dirname(out_clinical_data_file), exist_ok=True)
+selected_clinical.to_csv(out_clinical_data_file)
+
+# 2) Normalize gene symbols and save normalized gene expression
+# Ensure numeric gene expression
+genetic_df_numeric = genetic_df.apply(pd.to_numeric, errors='coerce')
+normalized_gene_df = normalize_gene_symbols_in_index(genetic_df_numeric.copy())
+os.makedirs(os.path.dirname(out_gene_data_file), exist_ok=True)
+normalized_gene_df.to_csv(out_gene_data_file)
+
+# 3) Link clinical and genetic data on sample IDs
+gene_expr_T = normalized_gene_df.T  # samples as rows, genes as columns
+linked_data = selected_clinical.join(gene_expr_T, how='inner')
+
+# 4) Handle missing values systematically
+processed_df = handle_missing_values(linked_data.copy(), trait_col=trait)
+
+# 5) Determine bias and remove biased demographic features if needed
+trait_biased, processed_df = judge_and_remove_biased_features(processed_df, trait=trait)
+
+# 6) Final validation and save cohort info
+# Force native Python bools to avoid JSON serialization issues
+is_gene_available = bool((normalized_gene_df.shape[0] > 0) and (normalized_gene_df.shape[1] > 0))
+is_trait_available = bool((trait in linked_data.columns) and bool(linked_data[trait].notna().any()))
+trait_biased_py = bool(trait_biased)
+
+note = "INFO: Age derived from days_to_birth and converted to years when available."
+is_usable = False
+try:
+    is_usable = validate_and_save_cohort_info(
+        is_final=True,
+        cohort="TCGA",
+        info_path=json_path,
+        is_gene_available=is_gene_available,
+        is_trait_available=is_trait_available,
+        is_biased=trait_biased_py,
+        df=processed_df,
+        note=note
+    )
+except TypeError:
+    # Retry once with a sanitized DataFrame to ensure standard metadata types
+    processed_df_safe = processed_df.copy()
+    processed_df_safe.columns = processed_df_safe.columns.map(str)
+    is_usable = validate_and_save_cohort_info(
+        is_final=True,
+        cohort="TCGA",
+        info_path=json_path,
+        is_gene_available=bool(is_gene_available),
+        is_trait_available=bool(is_trait_available),
+        is_biased=bool(trait_biased_py),
+        df=processed_df_safe,
+        note=note
+    )
+
+# 7) Save usable linked data
+if is_usable:
+    os.makedirs(os.path.dirname(out_data_file), exist_ok=True)
+    processed_df.to_csv(out_data_file)

output/preprocess/Thymoma/cohort_info.json

@@ -1,42 +1 @@
-{
-  "GSE42977": {
-    "is_usable": true,
-    "is_gene_available": true,
-    "is_trait_available": true,
-    "is_available": true,
-    "is_biased": false,
-    "has_age": false,
-    "has_gender": false,
-    "sample_size": 117
-  },
-  "GSE29695": {
-    "is_usable": false,
-    "is_gene_available": true,
-    "is_trait_available": true,
-    "is_available": true,
-    "is_biased": true,
-    "has_age": false,
-    "has_gender": false,
-    "sample_size": 41
-  },
-  "GSE131027": {
-    "is_usable": false,
-    "is_gene_available": true,
-    "is_trait_available": true,
-    "is_available": true,
-    "is_biased": true,
-    "has_age": false,
-    "has_gender": false,
-    "sample_size": 92
-  },
-  "TCGA": {
-    "is_usable": false,
-    "is_gene_available": true,
-    "is_trait_available": true,
-    "is_available": true,
-    "is_biased": true,
-    "has_age": true,
-    "has_gender": true,
-    "sample_size": 122
-  }
-}
+{"GSE42977": {"is_usable": true, "is_gene_available": true, "is_trait_available": true, "is_available": true, "is_biased": false, "has_age": false, "has_gender": false, "sample_size": 117, "note": "INFO: Gene IDs mapped from Illumina ILMN probes to HGNC symbols; trait derived from 'tissue' field where any mention of 'Thymoma' is labeled as case (1)."}, "GSE29695": {"is_usable": false, "is_gene_available": true, "is_trait_available": false, "is_available": false, "is_biased": null, "has_age": null, "has_gender": null, "sample_size": null, "note": "INFO: Trait data unavailable in this series; skipped linking, missing value handling, and bias checks."}, "GSE131027": {"is_usable": false, "is_gene_available": true, "is_trait_available": true, "is_available": true, "is_biased": true, "has_age": false, "has_gender": false, "sample_size": 92, "note": "WARNING: Severe class imbalance for 'Thymoma': 1 positive vs 91 negative (N=92)."}, "TCGA": {"is_usable": false, "is_gene_available": true, "is_trait_available": true, "is_available": true, "is_biased": true, "has_age": true, "has_gender": true, "sample_size": 122, "note": "INFO: Age derived from days_to_birth and converted to years when available."}}

output/preprocess/Thyroid_Cancer/clinical_data/GSE104006.csv

@@ -0,0 +1,4 @@
+,GSM2787513,GSM2787514,GSM2787515,GSM2787516,GSM2787517,GSM2787518,GSM2787519,GSM2787520,GSM2787521,GSM2787522,GSM2787523,GSM2787524,GSM2787525,GSM2787526,GSM2787527,GSM2787528,GSM2787529,GSM2787530,GSM2787531,GSM2787532,GSM2787533,GSM2787534,GSM2787535,GSM2787536,GSM2787537,GSM2787538,GSM2787539,GSM2787540,GSM2787541,GSM2787542,GSM2787543,GSM2787544,GSM2787545,GSM2787546
+Thyroid_Cancer,1.0,1.0,1.0,1.0,1.0,1.0,1.0,0.0,1.0,1.0,1.0,0.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,0.0,1.0,1.0,0.0,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,1.0,1.0
+Age,74.0,74.0,72.0,74.0,38.0,50.0,41.0,51.0,73.0,52.0,48.0,59.0,58.0,39.0,37.0,33.0,36.0,70.0,26.0,46.0,57.0,44.0,35.0,42.0,47.0,61.0,38.0,35.0,35.0,38.0,49.0,56.0,52.0,51.0
+Gender,1.0,1.0,0.0,0.0,0.0,0.0,1.0,0.0,1.0,0.0,0.0,1.0,1.0,0.0,0.0,0.0,1.0,0.0,0.0,1.0,0.0,0.0,1.0,1.0,0.0,0.0,0.0,1.0,0.0,0.0,1.0,1.0,0.0,0.0

output/preprocess/Thyroid_Cancer/clinical_data/GSE107754.csv

@@ -1,3 +1,3 @@
 ,GSM2878070,GSM2878071,GSM2878072,GSM2878073,GSM2878074,GSM2878075,GSM2878076,GSM2878077,GSM2878078,GSM2878079,GSM2878080,GSM2878081,GSM2878082,GSM2891194,GSM2891195,GSM2891196,GSM2891197,GSM2891198,GSM2891199,GSM2891200,GSM2891201,GSM2891202,GSM2891203,GSM2891204,GSM2891205,GSM2891206,GSM2891207,GSM2891208,GSM2891209,GSM2891210,GSM2891211,GSM2891212,GSM2891213,GSM2891214,GSM2891215,GSM2891216,GSM2891217,GSM2891218,GSM2891219,GSM2891220,GSM2891221,GSM2891222,GSM2891223,GSM2891224,GSM2891225,GSM2891226,GSM2891227,GSM2891228,GSM2891229,GSM2891230,GSM2891231,GSM2891232,GSM2891233,GSM2891234,GSM2891235,GSM2891236,GSM2891237,GSM2891238,GSM2891239,GSM2891240,GSM2891241,GSM2891242,GSM2891243,GSM2891244,GSM2891245,GSM2891246,GSM2891247,GSM2891248,GSM2891249,GSM2891250,GSM2891251,GSM2891252,GSM2891253,GSM2891254,GSM2891255,GSM2891256,GSM2891257,GSM2891258,GSM2891259,GSM2891260,GSM2891261,GSM2891262,GSM2891263,GSM2891264
-Thyroid_Cancer,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,1.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,1.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0
+Thyroid_Cancer,,,,,,,,,,,,,,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,1.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,1.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0
 Gender,1.0,0.0,1.0,1.0,0.0,1.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,1.0,1.0,0.0,0.0,0.0,0.0,0.0,1.0,1.0,0.0,0.0,1.0,0.0,0.0,1.0,1.0,0.0,0.0,0.0,0.0,0.0,1.0,0.0,1.0,0.0,0.0,1.0,0.0,1.0,0.0,1.0,1.0,0.0,1.0,0.0,0.0,1.0,1.0,1.0,1.0,0.0,0.0,1.0,0.0,0.0,0.0,1.0,1.0,1.0,0.0,0.0,0.0,0.0,1.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,1.0,1.0,1.0,1.0,0.0,1.0,0.0,1.0,1.0,1.0

output/preprocess/Thyroid_Cancer/clinical_data/GSE138198.csv

@@ -1,3 +1,3 @@
 ,GSM4101749,GSM4101750,GSM4101751,GSM4101752,GSM4101753,GSM4101754,GSM4101755,GSM4101756,GSM4101757,GSM4101758,GSM4101759,GSM4101760,GSM4101761,GSM4101762,GSM4101763,GSM4101764,GSM4101765,GSM4101766,GSM4101767,GSM4101768,GSM4101769,GSM4101770,GSM4101771,GSM4101772,GSM4101773,GSM4101774,GSM4101775,GSM4101776,GSM4101777,GSM4101778,GSM4101779,GSM4101780,GSM4101781,GSM4101782,GSM4101783,GSM4101784
-Thyroid_Cancer,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
+Thyroid_Cancer,,,,,,,,,,,,,,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0
 Gender,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,1.0,0.0,1.0,0.0,0.0,1.0,0.0,0.0,0.0,1.0,0.0,0.0,0.0,1.0,1.0,0.0,0.0,0.0,1.0,0.0,0.0,,,

output/preprocess/Thyroid_Cancer/clinical_data/GSE151179.csv

@@ -1,3 +1,2 @@
-,GSM4567912,GSM4567913,GSM4567914,GSM4567915,GSM4567916,GSM4567917,GSM4567918,GSM4567919,GSM4567920,GSM4567921,GSM4567922,GSM4567923,GSM4567924,GSM4567925,GSM4567926,GSM4567927,GSM4567928,GSM4567929,GSM4567930,GSM4567931,GSM4567932,GSM4567933,GSM4567934,GSM4567935,GSM4567936,GSM4567937,GSM4567938,GSM4567939,GSM4567940,GSM4567941,GSM4567942,GSM4567943,GSM4567944,GSM4567945,GSM4567946,GSM4567947,GSM4567948,GSM4567949,GSM4567950,GSM4567951,GSM4567952,GSM4567953,GSM4567954,GSM4567955,GSM4567956,GSM4567957,GSM4567958,GSM4567959,GSM4567960,GSM4567961,GSM4567962,GSM4567963
-Thyroid_Cancer,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0
-Gender,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
+GSM4567912,GSM4567913,GSM4567914,GSM4567915,GSM4567916,GSM4567917,GSM4567918,GSM4567919,GSM4567920,GSM4567921,GSM4567922,GSM4567923,GSM4567924,GSM4567925,GSM4567926,GSM4567927,GSM4567928,GSM4567929,GSM4567930,GSM4567931,GSM4567932,GSM4567933,GSM4567934,GSM4567935,GSM4567936,GSM4567937,GSM4567938,GSM4567939,GSM4567940,GSM4567941,GSM4567942,GSM4567943,GSM4567944,GSM4567945,GSM4567946,GSM4567947,GSM4567948,GSM4567949,GSM4567950,GSM4567951,GSM4567952,GSM4567953,GSM4567954,GSM4567955,GSM4567956,GSM4567957,GSM4567958,GSM4567959,GSM4567960,GSM4567961,GSM4567962,GSM4567963
+1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0

output/preprocess/Thyroid_Cancer/clinical_data/GSE151181.csv

@@ -1,2 +1,2 @@
-GSM4567912,GSM4567913,GSM4567914,GSM4567915,GSM4567916,GSM4567917,GSM4567918,GSM4567919,GSM4567920,GSM4567921,GSM4567922,GSM4567923,GSM4567924,GSM4567925,GSM4567926,GSM4567927,GSM4567928,GSM4567929,GSM4567930,GSM4567931,GSM4567932,GSM4567933,GSM4567934,GSM4567935,GSM4567936,GSM4567937,GSM4567938,GSM4567939,GSM4567940,GSM4567941,GSM4567942,GSM4567943,GSM4567944,GSM4567945,GSM4567946,GSM4567947,GSM4567948,GSM4567949,GSM4567950,GSM4567951,GSM4567952,GSM4567953,GSM4567954,GSM4567955,GSM4567956,GSM4567957,GSM4567958,GSM4567959,GSM4567960,GSM4567961,GSM4567962,GSM4567963
-0.0,0.0,0.0,0.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0
+,GSM4567912,GSM4567913,GSM4567914,GSM4567915,GSM4567916,GSM4567917,GSM4567918,GSM4567919,GSM4567920,GSM4567921,GSM4567922,GSM4567923,GSM4567924,GSM4567925,GSM4567926,GSM4567927,GSM4567928,GSM4567929,GSM4567930,GSM4567931,GSM4567932,GSM4567933,GSM4567934,GSM4567935,GSM4567936,GSM4567937,GSM4567938,GSM4567939,GSM4567940,GSM4567941,GSM4567942,GSM4567943,GSM4567944,GSM4567945,GSM4567946,GSM4567947,GSM4567948,GSM4567949,GSM4567950,GSM4567951,GSM4567952,GSM4567953,GSM4567954,GSM4567955,GSM4567956,GSM4567957,GSM4567958,GSM4567959,GSM4567960,GSM4567961,GSM4567962,GSM4567963
+Thyroid_Cancer,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0

output/preprocess/Thyroid_Cancer/clinical_data/GSE58689.csv

@@ -1,3 +1,4 @@
 ,GSM1413225,GSM1413226,GSM1413227,GSM1413228,GSM1413229,GSM1413230,GSM1413231,GSM1413232,GSM1413233,GSM1413234,GSM1413235,GSM1413236,GSM1413237,GSM1413238,GSM1413239,GSM1413240,GSM1413241,GSM1413242,GSM1413243,GSM1413244,GSM1413245,GSM1413246,GSM1413247,GSM1413248,GSM1413249,GSM1413250,GSM1413251,GSM1413252,GSM1413253,GSM1413254,GSM1413255,GSM1413256,GSM1413257,GSM1413258,GSM1413259,GSM1413260,GSM1413261,GSM1413262,GSM1413263,GSM1413264,GSM1413265,GSM1413266,GSM1413267,GSM1413268,GSM1413269
 Thyroid_Cancer,0.0,0.0,1.0,0.0,1.0,0.0,1.0,0.0,0.0,0.0,0.0,1.0,1.0,1.0,0.0,0.0,1.0,0.0,0.0,1.0,0.0,1.0,0.0,0.0,1.0,0.0,1.0,0.0,1.0,0.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0
-Gender,0.0,0.0,,0.0,,0.0,,1.0,0.0,0.0,0.0,,,,1.0,0.0,,0.0,1.0,,0.0,,0.0,0.0,,1.0,,0.0,,0.0,,,,,,,,,,,,,,,
+Age,,,47.0,,37.0,,67.0,,,,,26.0,32.0,61.0,,,28.0,,,12.0,,69.0,,,6.0,,24.0,,31.0,,71.0,20.0,66.0,21.0,71.0,44.0,31.0,28.0,44.0,30.0,21.0,19.0,59.0,64.0,23.0
+Gender,,,0.0,,0.0,,1.0,,,,,0.0,0.0,1.0,,,0.0,,,1.0,,0.0,,,0.0,,0.0,,0.0,,0.0,1.0,0.0,0.0,1.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,1.0

output/preprocess/Thyroid_Cancer/clinical_data/GSE80022.csv

@@ -1,4 +1,2 @@
 ,GSM2111207,GSM2111208,GSM2111209,GSM2111210,GSM2111211,GSM2111212,GSM2111213,GSM2111214,GSM2111215,GSM2111216,GSM2111217,GSM2111218,GSM2111219,GSM2111220,GSM2111221,GSM2111222,GSM2111223,GSM2111224,GSM2111225,GSM2111226,GSM2111227,GSM2111228,GSM2111229,GSM2111230,GSM2111231,GSM2111232,GSM2111233,GSM2111234,GSM2111235,GSM2111236,GSM2111237,GSM2111238,GSM2111239,GSM2111240,GSM2111241,GSM2111242,GSM2111243,GSM2111244,GSM2111245,GSM2111246,GSM2111247,GSM2111248,GSM2111249,GSM2111250,GSM2111251,GSM2111252,GSM2111253,GSM2111254,GSM2111255,GSM2111256,GSM2111257,GSM2111258,GSM2111259,GSM2111260,GSM2111261,GSM2111262,GSM2111263,GSM2111264,GSM2111265,GSM2111266,GSM2111267,GSM2111268,GSM2111269,GSM2111270,GSM2111271,GSM2111272,GSM2111273,GSM2111274,GSM2111275,GSM2111276,GSM2111277,GSM2111278,GSM2111279
-Thyroid_Cancer,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
-Age,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
-Gender,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
+Thyroid_Cancer,0.0,0.0,0.0,0.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0

output/preprocess/Thyroid_Cancer/code/GSE104005.py

@@ -0,0 +1,197 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Thyroid_Cancer"
+cohort = "GSE104005"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Thyroid_Cancer"
+in_cohort_dir = "../DATA/GEO/Thyroid_Cancer/GSE104005"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Thyroid_Cancer/GSE104005.csv"
+out_gene_data_file = "./output/z6/preprocess/Thyroid_Cancer/gene_data/GSE104005.csv"
+out_clinical_data_file = "./output/z6/preprocess/Thyroid_Cancer/clinical_data/GSE104005.csv"
+json_path = "./output/z6/preprocess/Thyroid_Cancer/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
+import re
+
+# 1) Gene expression availability
+is_gene_available = True  # Illumina HumanHT-12 expression beadchip mentioned in background
+
+# 2) Variable availability (from provided Sample Characteristics Dictionary)
+trait_row = 0   # 'disease: Thyroid_carcinoma' vs 'disease: Non-neoplastic_thyroid'
+age_row = 2     # 'age: <number>'
+gender_row = 3  # 'Sex: F' / 'Sex: M'
+
+# 2.2) Conversion functions
+def _after_colon(value):
+    if value is None:
+        return None
+    if isinstance(value, str):
+        parts = value.split(":", 1)
+        val = parts[1] if len(parts) > 1 else parts[0]
+        return val.strip()
+    return str(value).strip()
+
+def convert_trait(value):
+    v = _after_colon(value)
+    if v is None or v == '':
+        return None
+    norm = v.lower().replace('_', ' ').replace('-', ' ').strip()
+    # Controls
+    if 'non' in norm and 'neoplastic' in norm:
+        return 0
+    if norm in {'normal', 'control', 'benign'}:
+        return 0
+    # Cases
+    cancer_keywords = ['carcinoma', 'cancer', 'tumor', 'tumour', 'ptc', 'pdtc', 'atc', 'metastasis', 'metastases']
+    if any(k in norm for k in cancer_keywords):
+        return 1
+    # Fallback: specific exact match
+    if norm == 'thyroid carcinoma':
+        return 1
+    return None
+
+def convert_age(value):
+    v = _after_colon(value)
+    if v is None or v == '':
+        return None
+    v_low = v.lower()
+    if v_low in {'na', 'n/a', 'unknown', 'null'}:
+        return None
+    match = re.search(r'(\d+(?:\.\d+)?)', v_low)
+    if not match:
+        return None
+    num = float(match.group(1))
+    # Return int if whole number to keep data neat
+    return int(num) if num.is_integer() else num
+
+def convert_gender(value):
+    v = _after_colon(value)
+    if v is None or v == '':
+        return None
+    norm = v.strip().lower()
+    if norm in {'f', 'female', 'woman', 'girl'}:
+        return 0
+    if norm in {'m', 'male', 'man', 'boy'}:
+        return 1
+    return None
+
+# 3) Save metadata with initial filtering
+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
+)
+
+# 4) Clinical feature extraction (only if trait data 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
+    )
+    clinical_preview = preview_df(selected_clinical_df)
+    print("Clinical data preview:", clinical_preview)
+
+    os.makedirs(os.path.dirname(out_clinical_data_file), exist_ok=True)
+    selected_clinical_df.to_csv(out_clinical_data_file)
+
+# 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
+print("requires_gene_mapping = True")
+
+# 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
+# Decide columns for probe ID and gene symbol based on annotation preview
+probe_col = 'ID'  # Matches ILMN_* probe IDs seen in expression data
+symbol_col = 'Symbol' if 'Symbol' in gene_annotation.columns else 'ILMN_Gene'
+
+# Build probe-to-gene mapping dataframe
+mapping_df = get_gene_mapping(gene_annotation, prob_col=probe_col, gene_col=symbol_col)
+
+# Apply mapping to convert probe-level data to gene-level expression
+gene_data = apply_gene_mapping(expression_df=gene_data, mapping_df=mapping_df)
+
+# Step 7: Data Normalization and Linking
+import os
+
+# 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 demographic features
+is_trait_biased, unbiased_linked_data = judge_and_remove_biased_features(linked_data, trait)
+
+# 5. Final validation and save cohort info using actual availability flags
+is_gene_available_final = (normalized_gene_data.shape[0] > 0) and (normalized_gene_data.shape[1] > 0)
+is_trait_available_final = trait in selected_clinical_df.index
+
+note = "INFO: Illumina HumanHT-12 probe data mapped to gene symbols; trait from 'disease' field; miRNA platform present but gene chip used."
+is_usable = validate_and_save_cohort_info(
+    is_final=True,
+    cohort=cohort,
+    info_path=json_path,
+    is_gene_available=is_gene_available_final,
+    is_trait_available=is_trait_available_final,
+    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)

output/preprocess/Thyroid_Cancer/code/GSE104006.py

@@ -0,0 +1,262 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Thyroid_Cancer"
+cohort = "GSE104006"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Thyroid_Cancer"
+in_cohort_dir = "../DATA/GEO/Thyroid_Cancer/GSE104006"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Thyroid_Cancer/GSE104006.csv"
+out_gene_data_file = "./output/z6/preprocess/Thyroid_Cancer/gene_data/GSE104006.csv"
+out_clinical_data_file = "./output/z6/preprocess/Thyroid_Cancer/clinical_data/GSE104006.csv"
+json_path = "./output/z6/preprocess/Thyroid_Cancer/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
+import re
+import math
+
+# 1) Gene expression availability (SuperSeries includes gene expression profiling, not just miRNA)
+is_gene_available = True
+
+# 2) Variable availability based on the provided Sample Characteristics Dictionary
+trait_row = 0        # 'disease: Thyroid_carcinoma' vs 'disease: Non-neoplastic_thyroid'
+age_row = 2          # 'age: <number>'
+gender_row = 3       # 'Sex: F/M'
+
+# 2.2 Conversion functions
+def _after_colon(x):
+    if x is None:
+        return None
+    s = str(x).strip()
+    parts = s.split(':', 1)
+    val = parts[1] if len(parts) > 1 else parts[0]
+    return val.strip().strip('"').strip("'")
+
+def convert_trait(x):
+    v = _after_colon(x)
+    if v is None or v == '':
+        return None
+    vl = v.lower()
+    # Explicit mappings
+    if vl in {'thyroid_carcinoma'}:
+        return 1
+    if 'non-neoplastic' in vl or vl in {'non-neoplastic_thyroid', 'normal', 'control'}:
+        return 0
+    # Heuristic for robustness
+    if any(k in vl for k in ['carcinoma', 'cancer', 'tumor', 'malignant']):
+        return 1
+    if any(k in vl for k in ['benign', 'adjacent normal', 'healthy']):
+        return 0
+    return None
+
+def convert_age(x):
+    v = _after_colon(x)
+    if v is None or v == '':
+        return None
+    # Extract first number
+    m = re.search(r'[-+]?\d+(\.\d+)?', v)
+    if not m:
+        return None
+    try:
+        age = float(m.group())
+        if 0 <= age <= 120:
+            return age
+        return None
+    except Exception:
+        return None
+
+def convert_gender(x):
+    v = _after_colon(x)
+    if v is None or v == '':
+        return None
+    vl = v.lower()
+    if vl in {'f', 'female', 'women', 'woman'}:
+        return 0
+    if vl in {'m', 'male', 'men', 'man'}:
+        return 1
+    return None
+
+# 3) Save metadata with initial filtering
+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
+)
+
+# 4) Clinical Feature Extraction (only if clinical data 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
+    )
+    print(preview_df(selected_clinical_df))
+    os.makedirs(os.path.dirname(out_clinical_data_file), exist_ok=True)
+    selected_clinical_df.to_csv(out_clinical_data_file)
+
+# 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
+# Identifiers like 'hsa-let-7a-5p' are miRNA IDs (miRBase style), not HGNC gene symbols; thus mapping would be required.
+requires_gene_mapping = True
+print(f"requires_gene_mapping = {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
+# Robust mapping with platform consistency checks. Do not keep unmapped miRNA data as gene_data.
+
+import re
+
+# Preserve original expression data
+expr_df = gene_data.copy()
+expr_ids = pd.Index(expr_df.index.astype(str))
+
+# Heuristics: detect miRNA-like identifiers in expression data
+mirna_like = expr_ids.str.contains(r'(^hsa-)|(^miR)|(^mir-)|(-3p$)|(-5p$)', case=False, regex=True)
+mirna_fraction = mirna_like.mean()
+
+# Quick peek of annotation ID-like columns
+candidate_id_cols = [c for c in gene_annotation.columns if c.lower() in {'id', 'name', 'probe_id', 'probeid', 'transcript'}]
+if not candidate_id_cols:
+    candidate_id_cols = list(gene_annotation.columns)
+
+# Compute overlaps and show a few intersecting IDs to validate
+overlap_stats = {}
+overlap_examples = {}
+expr_id_set = set(expr_ids)
+for col in candidate_id_cols:
+    try:
+        ann_vals = gene_annotation[col].dropna().astype(str).str.strip()
+        ann_set = set(ann_vals)
+        inter = expr_id_set.intersection(ann_set)
+        overlap_stats[col] = len(inter)
+        if inter:
+            overlap_examples[col] = list(sorted(list(inter))[:5])
+    except Exception:
+        continue
+
+# Select the best ID column
+best_id_col = None
+best_overlap = -1
+for col, n in overlap_stats.items():
+    if n > best_overlap:
+        best_overlap = n
+        best_id_col = col
+
+total_ids = len(expr_ids)
+print(f"Detected miRNA-like identifiers in expression matrix: {mirna_fraction:.2%} of rows")
+print(f"Annotation-ID overlap by column (top 10): {dict(list(sorted(overlap_stats.items(), key=lambda x: -x[1]))[:10])}")
+if best_id_col is not None and best_overlap > 0:
+    print(f"Sample intersecting IDs for column '{best_id_col}': {overlap_examples.get(best_id_col, [])}")
+print(f"Selected ID column: {best_id_col!r} with {best_overlap} overlapping IDs out of {total_ids}")
+
+# Choose a gene symbol column with common names preference
+preferred_symbol_cols = [
+    'Symbol', 'Gene Symbol', 'Gene_Symbol', 'Gene symbol', 'Gene', 'GENE_SYMBOL',
+    'ILMN_Gene', 'HGNC_symbol', 'HGNC', 'miRNA_ID', 'miRNA', 'MIRNA_SYMBOL', 'Transcript'
+]
+gene_col = next((c for c in preferred_symbol_cols if c in gene_annotation.columns), None)
+if gene_col is None:
+    # Fallback to any column that is not the ID column
+    candidates = [c for c in gene_annotation.columns if c != best_id_col]
+    gene_col = candidates[0] if candidates else gene_annotation.columns[0]
+print(f"Selected gene symbol column: {gene_col!r}")
+
+# Platform consistency and mapping decision:
+platform_mismatch = False
+# If expression looks like miRNA while annotation IDs look like ILMN_* (mRNA), declare mismatch
+ann_id_sample = gene_annotation[best_id_col].dropna().astype(str).str.strip() if best_id_col is not None else pd.Series([], dtype=str)
+ann_looks_ilmn = ann_id_sample.str.startswith('ILMN_').mean() > 0.1 if len(ann_id_sample) > 0 else False
+
+if mirna_fraction > 0.5 and ann_looks_ilmn:
+    platform_mismatch = True
+
+# If there is effectively no overlap, also treat as mismatch
+minimal_overlap = best_overlap <= max(5, int(0.001 * total_ids))
+
+if platform_mismatch or minimal_overlap or best_id_col is None:
+    print("ERROR: Platform mismatch or insufficient overlap between expression IDs and annotation IDs detected.")
+    print("The matrix file appears to be miRNA/small RNA (e.g., 'hsa-let-7*'), while the SOFT annotation corresponds to an Illumina mRNA array (e.g., 'ILMN_*').")
+    print("Mapping to human gene symbols cannot be performed with these files. Marking gene data as unavailable for this cohort.")
+    # Do not keep miRNA-level data as gene_data; set to empty to fail downstream QC gracefully.
+    gene_data = pd.DataFrame()
+    try:
+        # Update initial metadata to reflect unavailability of gene-level data
+        _ = validate_and_save_cohort_info(
+            is_final=False,
+            cohort=cohort,
+            info_path=json_path,
+            is_gene_available=False,
+            is_trait_available=(True if 'trait_row' in globals() and trait_row is not None else False)
+        )
+    except Exception as e:
+        print(f"Metadata update warning (non-fatal): {e}")
+else:
+    # Build mapping and apply it
+    mapping_df = get_gene_mapping(gene_annotation, prob_col=best_id_col, gene_col=gene_col)
+    gene_data_mapped = apply_gene_mapping(expression_df=expr_df, mapping_df=mapping_df)
+
+    # Validate mapped result
+    if gene_data_mapped is None or gene_data_mapped.empty:
+        print("ERROR: Mapping produced an empty gene-level dataframe. Not preserving original miRNA/probe-level data.")
+        gene_data = pd.DataFrame()
+        try:
+            _ = validate_and_save_cohort_info(
+                is_final=False,
+                cohort=cohort,
+                info_path=json_path,
+                is_gene_available=False,
+                is_trait_available=(True if 'trait_row' in globals() and trait_row is not None else False)
+            )
+        except Exception as e:
+            print(f"Metadata update warning (non-fatal): {e}")
+    else:
+        gene_data = gene_data_mapped
+        print(f"Mapping succeeded. Gene-level dataframe shape: {gene_data.shape}")
+        # Show a few gene symbols
+        print(f"Example mapped genes: {list(gene_data.index[:10])}")

output/preprocess/Thyroid_Cancer/code/GSE107754.py

@@ -0,0 +1,200 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Thyroid_Cancer"
+cohort = "GSE107754"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Thyroid_Cancer"
+in_cohort_dir = "../DATA/GEO/Thyroid_Cancer/GSE107754"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Thyroid_Cancer/GSE107754.csv"
+out_gene_data_file = "./output/z6/preprocess/Thyroid_Cancer/gene_data/GSE107754.csv"
+out_clinical_data_file = "./output/z6/preprocess/Thyroid_Cancer/clinical_data/GSE107754.csv"
+json_path = "./output/z6/preprocess/Thyroid_Cancer/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
+import pandas as pd
+
+# 1) Gene expression availability
+is_gene_available = True  # Whole human genome gene expression microarrays per series description.
+
+# 2) Variable availability and converters
+
+# Trait: Thyroid_Cancer can be inferred from "tissue: Thyroid cancer" under key 2
+trait_row = 2
+
+# Age: not available in the provided characteristics
+age_row = None
+
+# Gender: available under key 0
+gender_row = 0
+
+def convert_trait(x):
+    if x is None:
+        return None
+    s = str(x).strip()
+    if s == "":
+        return None
+    parts = s.split(":", 1)
+    if len(parts) == 2:
+        header = parts[0].strip().lower()
+        value = parts[1].strip().lower()
+        # Only use records that specify tissue for trait inference
+        if "tissue" in header:
+            if "thyroid" in value:
+                return 1
+            # Any other specified tissue (non-thyroid) -> control (0)
+            # We assume all entries are cancers/tumors per dataset description.
+            return 0
+        else:
+            # Irrelevant headers like dataset, biopsy location, suvmean35, etc.
+            return None
+    else:
+        # Fallback: no colon present, attempt heuristic
+        s_low = s.lower()
+        if "thyroid" in s_low:
+            return 1
+        return None
+
+def convert_age(x):
+    # Age not available
+    return None
+
+def convert_gender(x):
+    if x is None:
+        return None
+    s = str(x).strip()
+    if s == "":
+        return None
+    parts = s.split(":", 1)
+    val = parts[1].strip().lower() if len(parts) == 2 else s.lower()
+    if val in ["female", "f", "woman", "women"]:
+        return 0
+    if val in ["male", "m", "man", "men"]:
+        return 1
+    return None
+
+# 3) Save metadata using initial filtering
+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
+)
+
+# 4) Clinical feature extraction (only if trait_row is not None)
+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("Clinical features preview:", preview)
+
+    os.makedirs(os.path.dirname(out_clinical_data_file), exist_ok=True)
+    selected_clinical_df.to_csv(out_clinical_data_file)
+
+# 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
+print("requires_gene_mapping = True")
+
+# 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
+# Identify columns for probe IDs and gene symbols in the annotation
+probe_col = 'ID'
+gene_symbol_col = 'GENE_SYMBOL'
+
+# Build mapping dataframe from annotation
+mapping_df = get_gene_mapping(gene_annotation, prob_col=probe_col, gene_col=gene_symbol_col)
+
+# Apply mapping to convert probe-level data to gene-level expression
+gene_data = apply_gene_mapping(expression_df=gene_data, mapping_df=mapping_df)
+
+# Step 7: Data Normalization and Linking
+import os
+import pandas as pd
+
+# 1. Normalize gene symbols and save normalized 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)
+
+# Ensure clinical data is available in this scope; load from disk if necessary
+try:
+    selected_clinical_df
+except NameError:
+    selected_clinical_df = pd.read_csv(out_clinical_data_file, index_col=0)
+
+# 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 check and removal of biased demographic features
+is_trait_biased, unbiased_linked_data = judge_and_remove_biased_features(linked_data, trait)
+
+# 5. Final quality validation and save cohort metadata
+is_gene_available = (normalized_gene_data.shape[0] > 0) and (normalized_gene_data.shape[1] > 0)
+is_trait_available = trait in linked_data.columns
+note = "INFO: Trait inferred from tissue; mixed metastatic tumor cohort; no age feature; gender included."
+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)

output/preprocess/Thyroid_Cancer/code/GSE138198.py

@@ -0,0 +1,207 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Thyroid_Cancer"
+cohort = "GSE138198"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Thyroid_Cancer"
+in_cohort_dir = "../DATA/GEO/Thyroid_Cancer/GSE138198"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Thyroid_Cancer/GSE138198.csv"
+out_gene_data_file = "./output/z6/preprocess/Thyroid_Cancer/gene_data/GSE138198.csv"
+out_clinical_data_file = "./output/z6/preprocess/Thyroid_Cancer/clinical_data/GSE138198.csv"
+json_path = "./output/z6/preprocess/Thyroid_Cancer/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
+import pandas as pd
+import re
+
+# 1) Gene expression data availability
+is_gene_available = True  # Affymetrix Human Gene 1.0 ST arrays -> gene expression
+
+# 2) Variable availability
+trait_row = 1  # sample type
+age_row = None  # not available in the characteristics
+gender_row = 0  # gender
+
+is_trait_available = trait_row is not None
+
+# 2.2) Converters
+def _extract_value(x):
+    if x is None:
+        return ""
+    s = str(x)
+    if ":" in s:
+        s = s.split(":", 1)[1]
+    return s.strip()
+
+def convert_trait(x):
+    v = _extract_value(x).lower()
+    # Exclude descriptive line listing all categories
+    if ('and three normal' in v) or ("hashimoto's thyroiditis" in v and 'normal thyroid' in v):
+        return None
+    # Cancer present (catch-all for any PTC mention)
+    if ('ptc' in v) or ('mptc' in v):
+        return 1
+    # Non-cancer categories
+    if ('normal thyroid' in v) or re.search(r'\btn\b', v):
+        return 0
+    if ('hashimoto' in v) or re.search(r'\bht\b', v):
+        return 0
+    return None
+
+def convert_age(x):
+    return None  # age not available in this dataset
+
+def convert_gender(x):
+    v = _extract_value(x).lower()
+    if v in {'f', 'female'}:
+        return 0
+    if v in {'m', 'male'}:
+        return 1
+    if 'not available' in v or v == '':
+        return None
+    return None
+
+# 3) Save metadata (initial filtering)
+_ = 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) Clinical feature extraction and save
+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
+    )
+    print("Selected clinical shape:", selected_clinical_df.shape)
+    preview = preview_df(selected_clinical_df, n=5)
+    print(preview)
+    # Save clinical features
+    clinical_dir = os.path.dirname(out_clinical_data_file)
+    os.makedirs(clinical_dir, exist_ok=True)
+    selected_clinical_df.to_csv(out_clinical_data_file)
+
+# 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
+# Based on the provided identifiers (e.g., '7892501'), these are probe/transcript IDs (e.g., Affymetrix),
+# not human gene symbols. Hence mapping to gene symbols is required.
+print("requires_gene_mapping = True")
+
+# 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
+# Decide the appropriate columns for mapping:
+# - Probe/ID column in gene_annotation matches the expression IDs: 'ID'
+# - Gene symbol information is embedded in the 'gene_assignment' column
+probe_col = 'ID'
+gene_symbol_col = 'gene_assignment'
+
+# 2. Build the mapping dataframe (ID -> Gene)
+mapping_df = get_gene_mapping(gene_annotation, prob_col=probe_col, gene_col=gene_symbol_col)
+
+# 3. Apply the mapping to convert probe-level data to gene-level expression
+gene_data = apply_gene_mapping(expression_df=gene_data, mapping_df=mapping_df)
+
+# Step 7: Data Normalization and Linking
+import os
+
+# 1. Normalize gene symbols and save gene expression 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
+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 demographic features
+is_trait_biased, unbiased_linked_data = judge_and_remove_biased_features(linked_data, trait)
+
+# Ensure native Python bools for JSON serialization
+is_trait_biased = bool(is_trait_biased)
+is_gene_available = bool((normalized_gene_data.shape[0] > 0) and (normalized_gene_data.shape[1] > 0))
+is_trait_available = bool((trait in linked_data.columns) and bool(linked_data[trait].notna().any()))
+
+# 5. Final validation and save cohort info
+note = ("INFO: Affymetrix Human Gene 1.0 ST platform; age not available; gender partially missing; "
+        "probe-to-gene mapping via SOFT gene_assignment with equal-split aggregation.")
+
+try:
+    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
+    )
+except TypeError:
+    # If serialization fails due to non-native bool types in an existing file, recreate and retry once.
+    if os.path.exists(json_path):
+        os.remove(json_path)
+    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)

output/preprocess/Thyroid_Cancer/code/GSE151179.py

@@ -0,0 +1,316 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Thyroid_Cancer"
+cohort = "GSE151179"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Thyroid_Cancer"
+in_cohort_dir = "../DATA/GEO/Thyroid_Cancer/GSE151179"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Thyroid_Cancer/GSE151179.csv"
+out_gene_data_file = "./output/z6/preprocess/Thyroid_Cancer/gene_data/GSE151179.csv"
+out_clinical_data_file = "./output/z6/preprocess/Thyroid_Cancer/clinical_data/GSE151179.csv"
+json_path = "./output/z6/preprocess/Thyroid_Cancer/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
+# Step 1: Determine gene expression availability
+is_gene_available = True  # Gene expression profiling by Thermo Fisher Human Clariom S Assay is described
+
+# Step 2: Variable availability and converters
+
+# Identify rows in the sample characteristics dictionary
+trait_row = 1        # 'tissue type' differentiates tumor/metastasis vs non-neoplastic thyroid
+age_row = None       # No age info in the sample characteristics dictionary
+gender_row = None    # No gender info in the sample characteristics dictionary
+
+def _extract_value(x):
+    if x is None:
+        return None
+    if isinstance(x, str):
+        parts = x.split(":", 1)
+        val = parts[1] if len(parts) > 1 else parts[0]
+        val = val.strip()
+        return val if val != "" else None
+    return x
+
+def convert_trait(x):
+    v = _extract_value(x)
+    if v is None:
+        return None
+    v_low = v.lower()
+    # Map non-neoplastic thyroid to 0; any tumor/metastasis to 1
+    if "non-neoplastic" in v_low:
+        return 0
+    if ("tumor" in v_low) or ("metastasis" in v_low):
+        return 1
+    # Conservative fallback
+    return None
+
+def convert_age(x):
+    v = _extract_value(x)
+    if v is None:
+        return None
+    # Extract first numeric token as age
+    import re
+    m = re.search(r"(\d+(\.\d+)?)", str(v))
+    if not m:
+        return None
+    age = float(m.group(1))
+    if age <= 0 or age > 120:
+        return None
+    return age
+
+def convert_gender(x):
+    v = _extract_value(x)
+    if v is None:
+        return None
+    v_low = str(v).strip().lower()
+    if v_low in {"female", "f", "woman", "women"}:
+        return 0
+    if v_low in {"male", "m", "man", "men"}:
+        return 1
+    return None
+
+# Step 3: Save metadata (initial filtering)
+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 and save
+    preview = preview_df(selected_clinical_df)
+    print("Preview of selected clinical features:", preview)
+
+    os.makedirs(os.path.dirname(out_clinical_data_file), exist_ok=True)
+    selected_clinical_df.to_csv(out_clinical_data_file, index=False)
+
+# 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
+requires_gene_mapping = True
+print(f"requires_gene_mapping = {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
+# Improved mapping from probe/transcript IDs to clean human gene symbols
+
+import re
+import json
+
+# Expression IDs from the series matrix
+expr_ids = set(gene_data.index.astype(str))
+
+# 1) Identify the annotation column that matches expression IDs
+preferred_id_cols = [
+    'transcript_cluster_id', 'Transcript_Cluster_ID', 'transcript_cluster', 'cluster_id',
+    'Probe Set ID', 'PROBE_SET_ID', 'probeset_id', 'probeset', 'PROBESET_ID', 'probe_id', 'ProbeID',
+    'ID'
+]
+available_cols = list(gene_annotation.columns)
+
+seen = set()
+id_candidates = []
+for c in preferred_id_cols + list(available_cols):
+    if c in available_cols and c not in seen:
+        id_candidates.append(c)
+        seen.add(c)
+
+def compute_overlap(series, expr_ids):
+    try:
+        vals = series.astype(str)
+    except Exception:
+        return 0
+    return int(vals.isin(expr_ids).sum())
+
+overlap_counts = {c: compute_overlap(gene_annotation[c], expr_ids) for c in id_candidates}
+id_col = max(overlap_counts, key=overlap_counts.get)
+max_overlap = overlap_counts[id_col]
+
+total_expr_ids = len(expr_ids)
+min_required = max(50, int(0.005 * total_expr_ids))  # a bit looser than before but still meaningful
+
+if max_overlap < min_required:
+    print("Diagnostics: Overlap of annotation columns with expression IDs (top 20):")
+    for c, v in sorted(overlap_counts.items(), key=lambda x: x[1], reverse=True)[:20]:
+        print(f"  {c}: {v}")
+    print(f"Chosen id_col: {id_col} with overlap {max_overlap} out of {total_expr_ids} expression IDs.")
+    raise ValueError(
+        "Failed to find a suitable annotation ID column that matches the expression IDs."
+    )
+
+# 2) Choose a column containing gene symbols/descriptions
+# Prefer richer annotation text that includes canonical symbols in parentheses
+if 'SPOT_ID.1' in available_cols:
+    gene_col = 'SPOT_ID.1'
+elif 'SPOT_ID' in available_cols:
+    gene_col = 'SPOT_ID'
+else:
+    # fallback to any column with 'gene' in name, else the second column
+    candidates = [c for c in available_cols if re.search(r'gene|symbol|assign|annot', str(c), re.I)]
+    gene_col = candidates[0] if candidates else (available_cols[1] if len(available_cols) > 1 else available_cols[0])
+
+# 3) Parse clean canonical symbols from the chosen annotation column, filtering by synonym dictionary
+with open("./metadata/gene_synonym.json", "r") as f:
+    synonym_dict = json.load(f)
+valid_symbols = set(synonym_dict.keys())  # keys are uppercased synonyms
+
+gene_like_pattern = re.compile(r'^(?:[A-Z][A-Z0-9-]{1,9}|C\d+ORF\d+)$')
+
+def parse_symbols(text: str):
+    if not isinstance(text, str) or not text:
+        return []
+    tokens = []
+
+    # Extract tokens that appear inside parentheses, which often hold canonical symbols
+    # Limit inner content length to avoid capturing long descriptive phrases
+    paren_contents = re.findall(r'\(([^)]{1,50})\)', text)
+
+    for content in paren_contents:
+        # Split on common delimiters
+        parts = re.split(r'[;/,\s]+', content)
+        for p in parts:
+            pu = p.strip().upper()
+            if gene_like_pattern.match(pu) and pu in valid_symbols:
+                tokens.append(synonym_dict.get(pu, pu))
+
+    # Fallback: if none found from parentheses, use a conservative global scan then filter strictly
+    if not tokens:
+        from_candidates = extract_human_gene_symbols(text)
+        for c in from_candidates:
+            cu = c.strip().upper()
+            if gene_like_pattern.match(cu) and cu in valid_symbols:
+                tokens.append(synonym_dict.get(cu, cu))
+
+    # Deduplicate while preserving order
+    seen_tok = set()
+    cleaned = []
+    for t in tokens:
+        tu = t.strip().upper()
+        if tu and tu not in seen_tok:
+            cleaned.append(tu)
+            seen_tok.add(tu)
+    return cleaned
+
+# Build mapping: keep only rows with ID overlapping expression IDs and with at least one valid symbol
+ann_sub = gene_annotation[[id_col, gene_col]].dropna()
+ann_sub[id_col] = ann_sub[id_col].astype(str).str.strip()
+ann_sub = ann_sub[ann_sub[id_col].isin(expr_ids)].copy()
+
+parsed = ann_sub[gene_col].apply(parse_symbols)
+has_symbols_mask = parsed.apply(lambda lst: len(lst) > 0)
+ann_sub = ann_sub[has_symbols_mask].copy()
+parsed = parsed[has_symbols_mask]
+
+# Join parsed symbols into a space-separated string so that apply_gene_mapping will re-extract them cleanly
+ann_sub['Gene'] = parsed.apply(lambda lst: ' '.join(lst))
+ann_sub = ann_sub.rename(columns={id_col: 'ID'})
+
+# Sanity check coverage
+mapping_overlap = ann_sub['ID'].astype(str).isin(gene_data.index.astype(str)).sum()
+if mapping_overlap < min_required:
+    print(f"Diagnostics: mapping size={len(ann_sub)}, overlap with expression IDs={mapping_overlap}")
+    print(f"Selected id_col={id_col}, gene_col={gene_col}")
+    print("Example annotation IDs (first 10):", list(ann_sub['ID'].astype(str).head(10)))
+    print("Example expression IDs (first 10):", list(pd.Index(gene_data.index.astype(str)).unique()[:10]))
+    raise ValueError("Insufficient overlap between mapping IDs and expression IDs; aborting to avoid empty/erroneous gene data.")
+
+# 4) Apply mapping to obtain gene-level data
+mapping_df = ann_sub[['ID', 'Gene']]
+gene_data = apply_gene_mapping(expression_df=gene_data, mapping_df=mapping_df)
+
+# 5) Quick validation: print top gene symbols to verify plausibility
+print(f"Gene-level data shape: {gene_data.shape}")
+print("First 20 gene symbols:", list(gene_data.index[:20]))
+
+# Step 7: Data Normalization and Linking
+import os
+import pandas as pd
+
+# 1. Normalize gene symbols and save gene expression data
+os.makedirs(os.path.dirname(out_gene_data_file), exist_ok=True)
+normalized_gene_data = normalize_gene_symbols_in_index(gene_data)
+normalized_gene_data.to_csv(out_gene_data_file)
+
+# 2. Reload clinical data from disk and set the correct index
+clinical_mat = pd.read_csv(out_clinical_data_file)
+clinical_df = clinical_mat.copy()
+clinical_df.index = [trait]  # single row with the trait as the feature name
+
+# Link clinical and genetic data
+linked_data = geo_link_clinical_genetic_data(clinical_df, normalized_gene_data)
+
+# 3. Handle missing values
+linked_data = handle_missing_values(linked_data, trait)
+
+# 4. Judge bias and remove biased demographic features (if any)
+is_trait_biased, unbiased_linked_data = judge_and_remove_biased_features(linked_data, trait)
+
+# 5. Final validation and save cohort metadata
+note = ("INFO: Age and Gender unavailable in sample characteristics; trait derived from 'tissue type'. "
+        "Thermo Fisher Human Clariom S platform used; probe-to-gene mapping and symbol normalization applied.")
+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,
+    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)

output/preprocess/Thyroid_Cancer/code/GSE151181.py

@@ -0,0 +1,251 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Thyroid_Cancer"
+cohort = "GSE151181"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Thyroid_Cancer"
+in_cohort_dir = "../DATA/GEO/Thyroid_Cancer/GSE151181"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Thyroid_Cancer/GSE151181.csv"
+out_gene_data_file = "./output/z6/preprocess/Thyroid_Cancer/gene_data/GSE151181.csv"
+out_clinical_data_file = "./output/z6/preprocess/Thyroid_Cancer/clinical_data/GSE151181.csv"
+json_path = "./output/z6/preprocess/Thyroid_Cancer/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
+import re
+import pandas as pd
+
+# 1) Gene expression data availability (not pure miRNA or methylation)
+# Series title indicates both gene and miRNA expression; presence of CIBERSORT purity also suggests mRNA data.
+is_gene_available = True
+
+# 2) Variable availability and conversion
+# Trait: Thyroid_Cancer (cancer vs non-neoplastic control)
+trait_row = 1  # 'tissue type' distinguishes non-neoplastic thyroid vs tumor/metastasis
+age_row = None
+gender_row = None
+
+def _extract_value(x):
+    if pd.isna(x):
+        return None
+    s = str(x)
+    # Extract substring after the first colon if present
+    parts = s.split(':', 1)
+    val = parts[1] if len(parts) > 1 else parts[0]
+    val = val.strip()
+    return val if val != '' else None
+
+def convert_trait(x):
+    v = _extract_value(x)
+    if v is None:
+        return None
+    vl = v.lower()
+    # Map non-cancer controls to 0
+    non_cancer_terms = [
+        'non-neoplastic thyroid', 'normal', 'adjacent normal', 'benign', 'normal thyroid'
+    ]
+    if vl in non_cancer_terms:
+        return 0
+    if vl in {'na', 'n/a', 'unknown', 'not available'}:
+        return None
+    # Any tumor/metastasis considered cancer = 1
+    cancer_indicators = ['tumor', 'metastasis', 'carcinoma', 'ptc', 'primary']
+    if any(term in vl for term in cancer_indicators):
+        return 1
+    return None
+
+def convert_age(x):
+    v = _extract_value(x)
+    if v is None:
+        return None
+    # find a number (integer or float)
+    m = re.search(r'(\d+(?:\.\d+)?)', v)
+    if not m:
+        return None
+    try:
+        age_val = float(m.group(1))
+    except Exception:
+        return None
+    if 0 < age_val < 120:
+        return age_val
+    return None
+
+def convert_gender(x):
+    v = _extract_value(x)
+    if v is None:
+        return None
+    vl = v.strip().lower()
+    if vl in {'female', 'f'}:
+        return 0
+    if vl in {'male', 'm'}:
+        return 1
+    return None
+
+# 3) Save metadata (initial filtering)
+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
+)
+
+# 4) Clinical feature extraction
+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
+    )
+    preview = preview_df(selected_clinical_df, n=5)
+    print(preview)
+
+    os.makedirs(os.path.dirname(out_clinical_data_file), exist_ok=True)
+    selected_clinical_df.to_csv(out_clinical_data_file)
+
+# 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
+# Based on observed gene identifiers from the previous step
+observed_ids = ['23064070', '23064071', '23064072', '23064073', '23064074',
+                '23064075', '23064076', '23064077', '23064078', '23064079',
+                '23064080', '23064081', '23064083', '23064084', '23064085',
+                '23064086', '23064087', '23064088', '23064089', '23064090']
+
+# If identifiers are purely numeric probe IDs, they need mapping to human gene symbols.
+requires_gene_mapping = all(x.isdigit() for x in observed_ids)
+
+print(f"requires_gene_mapping = {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
+import re
+import pandas as pd
+
+# 1) Decide which annotation column matches the expression identifiers and which holds gene symbols
+expr_ids = set(gene_data.index.astype(str))
+
+# Identify the best matching ID column in annotation (direct match or via extracting digits)
+best_match = {'col': None, 'method': None, 'matches': 0, 'series': None}
+for col in gene_annotation.columns:
+    ser = gene_annotation[col].astype(str).str.strip()
+
+    # Direct match
+    direct_matches = ser.isin(expr_ids).sum()
+    if direct_matches > best_match['matches']:
+        best_match = {'col': col, 'method': 'direct', 'matches': direct_matches, 'series': ser}
+
+    # Match after extracting a long digit token (e.g., from "ILMN_23064070" -> "23064070")
+    ser_digits = ser.str.extract(r'(\d{5,})', expand=False)
+    if ser_digits is not None:
+        digit_matches = ser_digits.isin(expr_ids).sum()
+        if digit_matches > best_match['matches']:
+            best_match = {'col': col, 'method': 'digits', 'matches': digit_matches, 'series': ser_digits}
+
+if best_match['matches'] == 0 or best_match['col'] is None:
+    raise ValueError("Failed to find an annotation column that matches expression probe IDs.")
+
+id_col = best_match['col']
+id_series = best_match['series']  # already processed per best method (either original or digits extracted)
+
+# Identify the best gene symbol column
+candidate_symbol_cols = [c for c in gene_annotation.columns if ('SYMBOL' in c.upper()) or ('GENE' in c.upper() and 'NAME' not in c.upper())]
+if not candidate_symbol_cols:
+    # Fallback to any column that might contain symbols
+    candidate_symbol_cols = list(gene_annotation.columns)
+
+def score_symbol_column(series: pd.Series) -> int:
+    def has_symbol(x):
+        syms = extract_human_gene_symbols(x)
+        return 1 if len(syms) > 0 else 0
+    return series.astype(str).apply(has_symbol).sum()
+
+symbol_scores = {c: score_symbol_column(gene_annotation[c]) for c in candidate_symbol_cols}
+gene_symbol_col = max(symbol_scores, key=symbol_scores.get)
+
+# 2) Build mapping dataframe from the chosen columns
+mapping_df = pd.DataFrame({
+    'ID': id_series.astype(str).str.strip(),
+    'Gene': gene_annotation[gene_symbol_col].astype(str)
+})
+
+# Clean mapping
+mapping_df = mapping_df.dropna(subset=['ID', 'Gene'])
+mapping_df = mapping_df[mapping_df['ID'].isin(expr_ids)]
+
+# 3) Apply mapping to convert probe-level to gene-level data
+gene_data = apply_gene_mapping(expression_df=gene_data, mapping_df=mapping_df)
+
+# Step 7: Data Normalization and Linking
+import os
+
+# 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 the clinical and genetic data (use the correct clinical variable name)
+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 demographic features are severely biased, and remove biased features
+is_trait_biased, unbiased_linked_data = judge_and_remove_biased_features(linked_data, trait)
+
+# 5. Final quality validation and save cohort information
+note = "INFO: Trait derived from 'tissue type' (non-neoplastic thyroid = 0, tumor/metastasis = 1)."
+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,
+    note=note
+)
+
+# 6. Save the 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)

output/preprocess/Thyroid_Cancer/code/GSE58689.py

@@ -0,0 +1,187 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Thyroid_Cancer"
+cohort = "GSE58689"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Thyroid_Cancer"
+in_cohort_dir = "../DATA/GEO/Thyroid_Cancer/GSE58689"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Thyroid_Cancer/GSE58689.csv"
+out_gene_data_file = "./output/z6/preprocess/Thyroid_Cancer/gene_data/GSE58689.csv"
+out_clinical_data_file = "./output/z6/preprocess/Thyroid_Cancer/clinical_data/GSE58689.csv"
+json_path = "./output/z6/preprocess/Thyroid_Cancer/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
+import re
+
+# 1) Gene expression data availability
+is_gene_available = True  # Based on series title indicating gene expression signature
+
+# 2) Variable availability and converters
+
+# Selected rows from the Sample Characteristics Dictionary
+trait_row = 0   # 'type of thyroid tissue: papillary thyroid carcinoma' vs 'normal thyroid'
+age_row = 3     # Exclusively age values listed
+gender_row = 2  # Contains 'Sex: female/male' among other entries; we'll parse only Sex and ignore others
+
+def _after_colon(val: str) -> str:
+    if val is None:
+        return ''
+    parts = str(val).split(':', 1)
+    return parts[1].strip() if len(parts) > 1 else str(val).strip()
+
+def convert_trait(x):
+    v = _after_colon(x).lower()
+    # Map thyroid tissue type to binary cancer status
+    if 'papillary' in v and 'carcinoma' in v:
+        return 1
+    if 'normal' in v:
+        return 0
+    # Heuristic for common abbreviations
+    if v in {'ptc', 'papillary thyroid carcinoma (ptc)'}:
+        return 1
+    return None
+
+def convert_age(x):
+    v = _after_colon(x)
+    # Extract the first number (integer or float)
+    m = re.search(r'[-+]?\d+\.?\d*', v)
+    if m:
+        try:
+            age_val = float(m.group())
+            # Basic sanity check for human age
+            if 0 <= age_val <= 120:
+                return age_val
+        except:
+            return None
+    return None
+
+def convert_gender(x):
+    s = str(x).lower()
+    # Prefer extraction after colon if present
+    v = _after_colon(s)
+    if 'female' in v:
+        return 0
+    if 'male' in v:
+        return 1
+    # If not in the value portion, try the whole string as a fallback
+    if 'female' in s:
+        return 0
+    if 'male' in s:
+        return 1
+    return None
+
+# 3) Save metadata (initial filtering)
+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
+)
+
+# 4) Clinical Feature Extraction
+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)
+
+    os.makedirs(os.path.dirname(out_clinical_data_file), exist_ok=True)
+    selected_clinical_df.to_csv(out_clinical_data_file)
+
+# 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
+# Affymetrix probe set IDs (e.g., "1007_s_at", "1053_at") are not human gene symbols and require mapping.
+requires_gene_mapping = True
+print(f"requires_gene_mapping = {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
+# Map Affymetrix probe set IDs ('ID') to human gene symbols ('Gene Symbol') and convert to gene-level matrix
+mapping_df = get_gene_mapping(gene_annotation, prob_col='ID', gene_col='Gene Symbol')
+gene_data = apply_gene_mapping(gene_data, mapping_df)
+
+# Step 7: Data Normalization and Linking
+import os
+
+# 1. Normalize gene symbols and save gene expression 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
+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. Evaluate bias and remove biased demographic features
+is_trait_biased, unbiased_linked_data = judge_and_remove_biased_features(linked_data, trait)
+
+# 5. Final validation and save cohort info
+is_gene_available = bool((normalized_gene_data.shape[0] > 0) and (normalized_gene_data.shape[1] > 0))
+
+has_trait_col = trait in unbiased_linked_data.columns
+has_trait_vals = bool(unbiased_linked_data[trait].notna().any()) if has_trait_col else False
+is_trait_available = bool(has_trait_col and has_trait_vals)
+
+note = (
+    "INFO: Affymetrix probe IDs mapped via 'Gene Symbol' from platform SOFT; "
+    "many samples lack explicit age/gender; missing values handled per protocol."
+)
+is_usable = validate_and_save_cohort_info(
+    True, cohort, json_path, is_gene_available, is_trait_available, is_trait_biased, unbiased_linked_data, note
+)
+
+# 6. Save linked dataset if usable
+if is_usable:
+    os.makedirs(os.path.dirname(out_data_file), exist_ok=True)
+    unbiased_linked_data.to_csv(out_data_file)

output/preprocess/Thyroid_Cancer/code/GSE76039.py

@@ -0,0 +1,182 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Thyroid_Cancer"
+cohort = "GSE76039"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Thyroid_Cancer"
+in_cohort_dir = "../DATA/GEO/Thyroid_Cancer/GSE76039"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Thyroid_Cancer/GSE76039.csv"
+out_gene_data_file = "./output/z6/preprocess/Thyroid_Cancer/gene_data/GSE76039.csv"
+out_clinical_data_file = "./output/z6/preprocess/Thyroid_Cancer/clinical_data/GSE76039.csv"
+json_path = "./output/z6/preprocess/Thyroid_Cancer/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
+# Step 1: Determine gene expression data availability based on background info
+is_gene_available = True  # Affymetrix U133 Plus 2.0 array indicates mRNA expression data
+
+# Step 2: Determine variable availability from the provided Sample Characteristics Dictionary
+# Sample Characteristics Dictionary:
+# {0: ['gender: female', 'gender: male'],
+#  1: ['tissue: Thyroid'],
+#  2: ['tumor type: Primary', 'tumor type: Recurrent tumor in neck',
+#      'tumor type: Recurrent/persistent metastasic tumor to lymph node',
+#      'tumor type: Metastasis', 'tumor type: Recurrent tumor', 'tumor type: Primary (residual)']}
+
+# Trait is "Thyroid_Cancer". All samples are tumor specimens; no controls. Trait is effectively constant -> not available.
+trait_row = None
+
+# Age is not found in the dictionary -> not available.
+age_row = None
+
+# Gender is available at key 0.
+gender_row = 0
+
+# Step 2.2: Define conversion functions
+import re
+from typing import Optional
+
+def _after_colon(value: str) -> str:
+    if value is None:
+        return ""
+    parts = str(value).split(":", 1)
+    return parts[1].strip() if len(parts) == 2 else str(value).strip()
+
+def convert_trait(x) -> Optional[int]:
+    # Not used because trait_row is None. For robustness, map any tumor-related entry to 1.
+    v = _after_colon(x).lower()
+    if v in ("", "na", "n/a", "unknown", "none"):
+        return None
+    # If any indication of tumor/tissue thyroid, mark as 1 for Thyroid_Cancer context.
+    keywords = ["tumor", "thyroid", "pdtc", "atc", "anaplastic", "poorly-differentiated", "metastasis", "primary"]
+    if any(k in v for k in keywords):
+        return 1
+    return None
+
+def convert_age(x) -> Optional[float]:
+    # Not used because age_row is None. Robust parser for potential age strings.
+    v = _after_colon(x).lower()
+    if v in ("", "na", "n/a", "unknown", "none"):
+        return None
+    # Extract first number (integer or float)
+    m = re.search(r"(\d+(\.\d+)?)", v)
+    if not m:
+        return None
+    try:
+        return float(m.group(1))
+    except Exception:
+        return None
+
+def convert_gender(x) -> Optional[int]:
+    v = _after_colon(x).lower()
+    if v in ("", "na", "n/a", "unknown", "none"):
+        return None
+    if "female" in v or v == "f":
+        return 0
+    if "male" in v or v == "m":
+        return 1
+    return None
+
+# Step 3: Save metadata with initial filtering
+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 (skip because trait_row is None)
+# If trait_row were available, we would extract and save clinical features as below:
+if trait_row is not None:
+    selected_clin_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_df(selected_clin_df)
+    os.makedirs(os.path.dirname(out_clinical_data_file), exist_ok=True)
+    selected_clin_df.to_csv(out_clinical_data_file)
+
+# 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
+# Based on the observed identifiers (e.g., '1007_s_at', '1053_at'), these are Affymetrix probe set IDs, not human gene symbols.
+print("requires_gene_mapping = True")
+
+# 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
+# Determine appropriate columns for mapping based on annotation preview:
+# - Probe identifiers: 'ID' (e.g., '1007_s_at', matching matrix row IDs)
+# - Gene symbols: 'Gene Symbol'
+probe_col = 'ID'
+gene_symbol_col = 'Gene Symbol'
+
+# Build mapping dataframe (ID -> Gene)
+mapping_df = get_gene_mapping(gene_annotation, prob_col=probe_col, gene_col=gene_symbol_col)
+
+# Apply mapping to convert probe-level expression to gene-level expression
+gene_data = apply_gene_mapping(expression_df=gene_data, mapping_df=mapping_df)
+
+# Step 7: Data Normalization and Linking
+import os
+
+# 1. Normalize gene symbols and save gene expression 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. Since trait data is unavailable (from Step 2: trait_row is None), skip linking and downstream steps
+linked_data = None
+
+# 5. Record metadata (initial filtering) to reflect unavailable trait data and available gene data
+#    Do not attempt final validation without clinical/trait data.
+_ = validate_and_save_cohort_info(
+    is_final=False,
+    cohort=cohort,
+    info_path=json_path,
+    is_gene_available=True,
+    is_trait_available=False
+)

output/preprocess/Thyroid_Cancer/code/GSE80022.py

@@ -0,0 +1,190 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Thyroid_Cancer"
+cohort = "GSE80022"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Thyroid_Cancer"
+in_cohort_dir = "../DATA/GEO/Thyroid_Cancer/GSE80022"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Thyroid_Cancer/GSE80022.csv"
+out_gene_data_file = "./output/z6/preprocess/Thyroid_Cancer/gene_data/GSE80022.csv"
+out_clinical_data_file = "./output/z6/preprocess/Thyroid_Cancer/clinical_data/GSE80022.csv"
+json_path = "./output/z6/preprocess/Thyroid_Cancer/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
+import math
+import pandas as pd
+
+# 1) Gene expression availability (Transcriptomic profiling of xenografts => gene expression data present)
+is_gene_available = True
+
+# 2) Variable availability and conversion functions
+
+# Trait: Thyroid_Cancer can be inferred from xenograft tissue (GOT2 = medullary thyroid carcinoma; GOT1 = small intestine NET)
+trait_row = 1  # 'xenograft tissue: GOT1' / 'xenograft tissue: GOT2'
+
+# No human age/gender in xenograft mouse model metadata
+age_row = None
+gender_row = None
+
+def _extract_value(x):
+    if x is None or (isinstance(x, float) and math.isnan(x)):
+        return None
+    s = str(x)
+    parts = s.split(":", 1)
+    v = parts[1] if len(parts) == 2 else parts[0]
+    return v.strip()
+
+def convert_trait(x):
+    v = _extract_value(x)
+    if v is None:
+        return None
+    vl = v.lower()
+    # Map medullary thyroid carcinoma model (GOT2) to 1, GOT1 (small intestine NET) to 0
+    if "got2" in vl or "medullary" in vl or "thyroid" in vl:
+        return 1
+    if "got1" in vl or "small intestine" in vl or "neuroendocrine" in vl or "net" in vl:
+        return 0
+    return None
+
+def convert_age(x):
+    v = _extract_value(x)
+    if v is None:
+        return None
+    # Try to parse numeric age if present; otherwise return None
+    try:
+        # Remove common units if any
+        vv = "".join(ch for ch in v if (ch.isdigit() or ch in ".-"))
+        return float(vv) if vv not in ("", "-", ".", "-.") else None
+    except Exception:
+        return None
+
+def convert_gender(x):
+    v = _extract_value(x)
+    if v is None:
+        return None
+    vl = v.strip().lower()
+    if vl in ["male", "m", "man"]:
+        return 1
+    if vl in ["female", "f", "woman"]:
+        return 0
+    return None
+
+# 3) Save metadata (initial filtering)
+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
+)
+
+# 4) Clinical feature extraction and save
+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)
+
+    os.makedirs(os.path.dirname(out_clinical_data_file), exist_ok=True)
+    selected_clinical_df.to_csv(out_clinical_data_file)
+
+# 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
+print("requires_gene_mapping = True")
+
+# 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
+# Decide identifier and gene symbol columns based on annotation preview:
+# - Probe IDs match 'ILMN_...' in the 'ID' column.
+# - Gene symbols are in the 'Symbol' column.
+
+# 1-2) Build mapping dataframe
+mapping_df = get_gene_mapping(gene_annotation, prob_col='ID', gene_col='Symbol')
+
+# 3) Apply mapping to convert probe-level data to gene-level data
+expression_df = gene_data  # preserve the original probe-level data
+gene_data = apply_gene_mapping(expression_df, mapping_df)
+
+# Step 7: Data Normalization and Linking
+import os
+
+# 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. Assess bias and remove biased demographic features
+is_trait_biased, unbiased_linked_data = judge_and_remove_biased_features(linked_data, trait)
+
+# 5. Final validation and save cohort info
+note = ("INFO: Xenograft mouse model (GOT1 small intestine NET vs GOT2 medullary thyroid carcinoma); "
+        "mouse hosts (Balb/c nude), no human age/gender covariates available.")
+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,
+    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)

output/preprocess/Thyroid_Cancer/code/GSE82208.py

@@ -0,0 +1,226 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Thyroid_Cancer"
+cohort = "GSE82208"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Thyroid_Cancer"
+in_cohort_dir = "../DATA/GEO/Thyroid_Cancer/GSE82208"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Thyroid_Cancer/GSE82208.csv"
+out_gene_data_file = "./output/z6/preprocess/Thyroid_Cancer/gene_data/GSE82208.csv"
+out_clinical_data_file = "./output/z6/preprocess/Thyroid_Cancer/clinical_data/GSE82208.csv"
+json_path = "./output/z6/preprocess/Thyroid_Cancer/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
+import pandas as pd
+
+# 1. Gene expression data availability
+is_gene_available = True  # mRNA gene expression profiling per series description
+
+# 2. Variable availability (rows from Sample Characteristics Dictionary)
+trait_row = 2     # 'class: FTC' vs 'class: FTA'
+age_row = 1       # 'age (years): ...'
+gender_row = 0    # 'Sex: Female/Male/-'
+
+# 2.2 Conversion functions
+def _after_colon(val):
+    if val is None:
+        return None
+    s = str(val)
+    if ':' in s:
+        s = s.split(':', 1)[1]
+    return s.strip()
+
+def convert_trait(val):
+    v = _after_colon(val)
+    if v is None or v == '' or v == '-' or v.lower() == 'unknown':
+        return None
+    v_low = v.lower()
+    if 'ftc' in v_low:
+        return 1  # cancer
+    if 'fta' in v_low:
+        return 0  # benign
+    return None
+
+def convert_age(val):
+    v = _after_colon(val)
+    if v is None:
+        return None
+    v = v.strip()
+    if v in ['', '-', 'na', 'n/a', 'null', 'none']:
+        return None
+    try:
+        return float(v)
+    except Exception:
+        # Try to extract leading number if present
+        num = ''.join(ch for ch in v if (ch.isdigit() or ch == '.' or ch == '-'))
+        try:
+            return float(num) if num not in ['', '-', '.', '--'] else None
+        except Exception:
+            return None
+
+def convert_gender(val):
+    v = _after_colon(val)
+    if v is None:
+        return None
+    v_low = v.strip().lower()
+    if v_low in ['female', 'f']:
+        return 0
+    if v_low in ['male', 'm']:
+        return 1
+    return None
+
+# 3. Save metadata (initial filtering)
+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
+)
+
+# 4. Clinical feature extraction (only if clinical 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, n=5)
+    print(preview)
+
+    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
+# Affymetrix probe set IDs (e.g., 1007_s_at) are not human gene symbols and require mapping.
+requires_gene_mapping = True
+print(f"requires_gene_mapping = {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
+# Determine the appropriate columns for mapping: probe IDs ('ID') to gene symbols ('Gene Symbol')
+mapping_df = get_gene_mapping(gene_annotation, prob_col='ID', gene_col='Gene Symbol')
+
+# Apply mapping to convert probe-level data to gene-level expression
+gene_data = apply_gene_mapping(expression_df=gene_data, mapping_df=mapping_df)
+
+# Step 7: Data Normalization and Linking
+import os
+
+# 1. Normalize gene symbols and save
+used_fallback = False
+try:
+    normalized_gene_data = normalize_gene_symbols_in_index(gene_data)
+except Exception as e:
+    # Fallback if synonym file is missing or normalization fails
+    used_fallback = True
+    normalized_gene_data = gene_data.copy()
+    normalized_gene_data.index = normalized_gene_data.index.str.upper()
+    normalized_gene_data = normalized_gene_data[normalized_gene_data.index.notnull()]
+    normalized_gene_data = normalized_gene_data.groupby(normalized_gene_data.index).mean()
+
+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
+try:
+    selected_clinical_df  # noqa: F401
+except NameError:
+    # Load from disk if not present in current context
+    selected_clinical_df = pd.read_csv(out_clinical_data_file, index_col=0)
+
+linked_data = geo_link_clinical_genetic_data(selected_clinical_df, normalized_gene_data)
+
+# Record pre-missing and shape stats for notes
+pre_n_samples, pre_n_cols = linked_data.shape
+covariate_cols = [trait, 'Age', 'Gender']
+pre_gene_cols = [c for c in linked_data.columns if c not in covariate_cols]
+pre_n_genes = len(pre_gene_cols)
+pre_missing_age = int(linked_data['Age'].isna().sum()) if 'Age' in linked_data.columns else None
+pre_missing_gender = int(linked_data['Gender'].isna().sum()) if 'Gender' in linked_data.columns else None
+
+# 3. Handle missing values
+linked_data = handle_missing_values(linked_data, trait)
+
+# 4. Bias checks and remove biased covariates
+is_trait_biased, unbiased_linked_data = judge_and_remove_biased_features(linked_data, trait)
+
+# Prepare note
+post_n_samples, post_n_cols = unbiased_linked_data.shape
+post_gene_cols = [c for c in unbiased_linked_data.columns if c not in [trait, 'Age', 'Gender']]
+post_n_genes = len(post_gene_cols)
+note_parts = []
+if used_fallback:
+    note_parts.append("WARNING: Fallback normalization used (gene synonym map unavailable).")
+else:
+    note_parts.append("INFO: Gene symbols normalized using NCBI synonym map.")
+note_parts.append(f"INFO: Samples before/after missing-value handling: {pre_n_samples}/{post_n_samples}.")
+note_parts.append(f"INFO: Gene features before/after filtering: {pre_n_genes}/{post_n_genes}.")
+if pre_missing_age is not None:
+    note_parts.append(f"INFO: Missing Age before handling: {pre_missing_age}.")
+if pre_missing_gender is not None:
+    note_parts.append(f"INFO: Missing Gender before handling: {pre_missing_gender}.")
+note = " ".join(note_parts)
+
+# 5. Final quality validation and cohort info saving
+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,
+    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)

output/preprocess/Thyroid_Cancer/code/TCGA.py

@@ -0,0 +1,310 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Thyroid_Cancer"
+
+# Input paths
+tcga_root_dir = "../DATA/TCGA"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Thyroid_Cancer/TCGA.csv"
+out_gene_data_file = "./output/z6/preprocess/Thyroid_Cancer/gene_data/TCGA.csv"
+out_clinical_data_file = "./output/z6/preprocess/Thyroid_Cancer/clinical_data/TCGA.csv"
+json_path = "./output/z6/preprocess/Thyroid_Cancer/cohort_info.json"
+
+
+# Step 1: Initial Data Loading
+import os
+import pandas as pd
+
+# Step 1: Select the most relevant TCGA cohort directory for Thyroid Cancer
+subdirs = [d for d in os.listdir(tcga_root_dir) if os.path.isdir(os.path.join(tcga_root_dir, d))]
+selected_dirname = None
+for pat in ['Thyroid_Cancer', '(THCA)', 'Thyroid', 'THCA']:
+    matches = [d for d in subdirs if pat.lower() in d.lower()]
+    if matches:
+        # Choose the most specific (shortest match name as proxy)
+        selected_dirname = sorted(matches, key=lambda x: len(x))[0]
+        break
+
+if selected_dirname is None:
+    # No suitable directory found: record and stop further processing in this run
+    validate_and_save_cohort_info(
+        is_final=False,
+        cohort="TCGA",
+        info_path=json_path,
+        is_gene_available=False,
+        is_trait_available=False
+    )
+else:
+    cohort_dir = os.path.join(tcga_root_dir, selected_dirname)
+
+    # Step 2: Identify clinical and genetic file paths
+    clinical_file_path, genetic_file_path = tcga_get_relevant_filepaths(cohort_dir)
+
+    # Step 3: Load both files
+    clinical_df = pd.read_csv(clinical_file_path, sep='\t', index_col=0, low_memory=False)
+    genetic_df = pd.read_csv(genetic_file_path, sep='\t', index_col=0, low_memory=False)
+
+    # Step 4: Print clinical column names
+    print(clinical_df.columns.tolist())
+
+# Step 2: Find Candidate Demographic Features
+import os
+import re
+import pandas as pd
+
+# Columns obtained from the previous step
+columns_from_prev_step = ['_INTEGRATION', '_PATIENT', '_cohort', '_primary_disease', '_primary_site', 'additional_pharmaceutical_therapy', 'additional_radiation_therapy', 'additional_surgery_locoregional_procedure', 'age_at_initial_pathologic_diagnosis', 'bcr_followup_barcode', 'bcr_patient_barcode', 'bcr_sample_barcode', 'braf_gene_genotyping_outcome_lab_results_text', 'days_to_additional_surgery_metastatic_procedure', 'days_to_birth', 'days_to_collection', 'days_to_death', 'days_to_initial_pathologic_diagnosis', 'days_to_last_followup', 'days_to_new_tumor_event_additional_surgery_procedure', 'days_to_new_tumor_event_after_initial_treatment', 'extrathyroid_carcinoma_present_extension_status', 'first_degree_relative_history_thyrd_glnd_crcnm_dgnss_rltnshp_typ', 'followup_case_report_form_submission_reason', 'form_completion_date', 'gender', 'genotype_analysis_performed_indicator', 'genotyping_results_gene_mutation_not_reported_reason', 'histologic_disease_progression_present_indicator', 'histologic_disease_progression_present_type', 'histological_type', 'histological_type_other', 'history_of_neoadjuvant_treatment', 'i_131_first_administered_dose', 'i_131_subsequent_administered_dose', 'i_131_total_administered_dose', 'i_131_total_administered_preparation_technique', 'icd_10', 'icd_o_3_histology', 'icd_o_3_site', 'informed_consent_verified', 'initial_weight', 'is_ffpe', 'lost_follow_up', 'lymph_node_examined_count', 'lymph_node_preoperative_assessment_diagnostic_imaging_type', 'lymph_node_preoperative_scan_indicator', 'metastatic_neoplasm_confirmed_diagnosis_method_name', 'metastatic_neoplasm_confirmed_diagnosis_method_text', 'metastatic_site', 'neoplasm_depth', 'new_neoplasm_confirmed_diagnosis_method_name', 'new_neoplasm_event_occurrence_anatomic_site', 'new_neoplasm_event_type', 'new_neoplasm_occurrence_anatomic_site_text', 'new_tumor_event_additional_surgery_procedure', 'new_tumor_event_after_initial_treatment', 'number_of_lymphnodes_positive_by_he', 'oct_embedded', 'other_dx', 'other_genotyping_outcome_lab_results_text', 'other_metastatic_site', 'pathologic_M', 'pathologic_N', 'pathologic_T', 'pathologic_stage', 'pathologic_stage', 'pathology_report_file_name', 'patient_id', 'patient_personal_medical_history_thyroid_gland_disorder_name', 'patient_personal_medical_history_thyroid_other_specify_text', 'person_lifetime_risk_radiation_exposure_indicator', 'person_neoplasm_cancer_status', 'post_surgical_procedure_assessment_thyroid_gland_carcinoma_stats', 'postoperative_rx_tx', 'primary_lymph_node_presentation_assessment', 'primary_neoplasm_focus_type', 'primary_thyroid_gland_neoplasm_location_anatomic_site', 'radiation_therapy', 'radiation_therapy_administered_dose_text', 'radiation_therapy_administered_preparation_technique_text', 'radiosensitizing_agent_administered_indicator', 'ras_family_gene_genotyping_outcome_lab_results_text', 'residual_tumor', 'ret_ptc_rearrangement_genotyping_outcome_lab_results_text', 'sample_type', 'sample_type_id', 'system_version', 'targeted_molecular_therapy', 'therapeutic_procedure_new_neoplasm_required_additional_thrpy_typ', 'thyroid_gland_carcinoma_involvement_regional_lymph_node_type', 'tissue_prospective_collection_indicator', 'tissue_retrospective_collection_indicator', 'tissue_source_site', 'tumor_tissue_site', 'vial_number', 'vital_status', 'year_of_initial_pathologic_diagnosis', '_GENOMIC_ID_TCGA_THCA_exp_HiSeqV2_percentile', '_GENOMIC_ID_TCGA_THCA_hMethyl450', '_GENOMIC_ID_TCGA_THCA_mutation_bcm_gene', '_GENOMIC_ID_TCGA_THCA_RPPA', '_GENOMIC_ID_TCGA_THCA_exp_HiSeqV2', '_GENOMIC_ID_TCGA_THCA_gistic2', '_GENOMIC_ID_TCGA_THCA_PDMRNAseqCNV', '_GENOMIC_ID_TCGA_THCA_mutation_bcgsc_gene', '_GENOMIC_ID_TCGA_THCA_mutation', '_GENOMIC_ID_TCGA_THCA_PDMRNAseq', '_GENOMIC_ID_TCGA_THCA_mutation_broad_gene', '_GENOMIC_ID_TCGA_THCA_exp_HiSeqV2_exon', '_GENOMIC_ID_TCGA_THCA_miRNA_HiSeq', '_GENOMIC_ID_TCGA_THCA_exp_HiSeqV2_PANCAN', '_GENOMIC_ID_data/public/TCGA/THCA/miRNA_HiSeq_gene', '_GENOMIC_ID_TCGA_THCA_gistic2thd']
+
+# Map lowercase to original for robust matching and original-case output
+lower_to_orig = {c.lower(): c for c in columns_from_prev_step}
+
+# Identify candidate age columns with stricter rules to avoid false positives (e.g., "stage", "agent")
+age_pattern = re.compile(r'(^|_)age($|_)', flags=re.I)
+candidate_age_cols = []
+
+for c in columns_from_prev_step:
+    cl = c.lower()
+    if age_pattern.search(cl):
+        candidate_age_cols.append(c)
+
+# Explicitly include known TCGA age-related aliases if present
+age_aliases = ['age_at_initial_pathologic_diagnosis', 'days_to_birth']
+for alias in age_aliases:
+    if alias in lower_to_orig and lower_to_orig[alias] not in candidate_age_cols:
+        candidate_age_cols.append(lower_to_orig[alias])
+
+# Identify candidate gender columns
+candidate_gender_cols = [c for c in columns_from_prev_step if ('gender' in c.lower() or 'sex' in c.lower())]
+
+# Print required lists in strict format
+print(f"candidate_age_cols = {candidate_age_cols}")
+print(f"candidate_gender_cols = {candidate_gender_cols}")
+
+# Helper to locate cohort directory
+def _find_cohort_dir(root_dir: str, keyword: str = 'THCA') -> str:
+    for entry in sorted(os.listdir(root_dir)):
+        full = os.path.join(root_dir, entry)
+        if os.path.isdir(full) and keyword.lower() in entry.lower():
+            return full
+    for r, dnames, _ in os.walk(root_dir):
+        for d in dnames:
+            if keyword.lower() in d.lower():
+                return os.path.join(r, d)
+    return ""
+
+# Robust reader for clinical matrix
+def _read_matrix_any(path: str) -> pd.DataFrame:
+    for sep in ['\t', ',', '|']:
+        try:
+            df = pd.read_csv(path, sep=sep, header=0, index_col=0, dtype=str)
+            if isinstance(df, pd.DataFrame) and df.shape[1] > 0:
+                return df
+        except Exception:
+            continue
+    try:
+        return pd.read_table(path, header=0, index_col=0, dtype=str)
+    except Exception:
+        return None
+
+# Load clinical data and preview candidate columns (dropping entirely empty columns from preview)
+clinical_df = None
+try:
+    cohort_dir = _find_cohort_dir(tcga_root_dir, keyword='THCA')
+    if cohort_dir:
+        clinical_fp, _ = tcga_get_relevant_filepaths(cohort_dir)
+        clinical_df = _read_matrix_any(clinical_fp)
+except Exception:
+    clinical_df = None
+
+if clinical_df is not None:
+    if candidate_age_cols:
+        age_cols_present = [c for c in candidate_age_cols if c in clinical_df.columns]
+        if age_cols_present:
+            age_df = clinical_df[age_cols_present]
+            age_df = age_df.loc[:, age_df.notna().any(axis=0)]
+            if age_df.shape[1] > 0:
+                age_preview = preview_df(age_df, n=5)
+                print(age_preview)
+    if candidate_gender_cols:
+        gender_cols_present = [c for c in candidate_gender_cols if c in clinical_df.columns]
+        if gender_cols_present:
+            gender_df = clinical_df[gender_cols_present]
+            gender_df = gender_df.loc[:, gender_df.notna().any(axis=0)]
+            if gender_df.shape[1] > 0:
+                gender_preview = preview_df(gender_df, n=5)
+                print(gender_preview)
+
+# Step 3: Select Demographic Features
+# Heuristic selection of demographic columns based on candidate previews and simple validity checks
+
+# Helper to locate the preview dictionaries created in previous steps, if available
+def _find_preview_dict(candidate_cols):
+    preview = None
+    for name, val in globals().items():
+        if isinstance(val, dict) and val:
+            # Ensure all keys are within the candidate columns and values are lists (first 5 samples)
+            if all(k in candidate_cols for k in val.keys()) and all(isinstance(v, list) for v in val.values()):
+                # Prefer the smallest dict that matches (more likely the preview)
+                if preview is None or len(val) < len(preview):
+                    preview = val
+    return preview
+
+# Initialize defaults
+age_col = None
+gender_col = None
+
+# Safeguards in case candidate lists are missing
+_candidate_age_cols = candidate_age_cols if 'candidate_age_cols' in globals() else []
+_candidate_gender_cols = candidate_gender_cols if 'candidate_gender_cols' in globals() else []
+
+# Try to retrieve preview dictionaries from previous step
+age_preview_dict = _find_preview_dict(_candidate_age_cols) if _candidate_age_cols else None
+gender_preview_dict = _find_preview_dict(_candidate_gender_cols) if _candidate_gender_cols else None
+
+# Select age column
+if _candidate_age_cols:
+    if age_preview_dict:
+        best_col = None
+        best_score = -1
+        for col, vals in age_preview_dict.items():
+            # Convert preview values using the provided helper
+            converted = [tcga_convert_age(v) for v in vals]
+            # Plausibility: count values that look like realistic ages
+            plausible = [v for v in converted if v is not None and 0 < v < 120]
+            score = len(plausible)
+            # Prefer explicitly age-like name in tie
+            if score > best_score or (score == best_score and best_col and 'age' in col.lower() and 'age' not in best_col.lower()):
+                best_col = col
+                best_score = score
+        age_col = best_col if best_col is not None and best_score > 0 else None
+    else:
+        # Fallback by common TCGA naming
+        for preferred in ['age_at_initial_pathologic_diagnosis', 'age', 'age_at_diagnosis']:
+            if preferred in _candidate_age_cols:
+                age_col = preferred
+                break
+
+# If we can assess missingness on the full clinical_df, ensure no large proportion missing
+if age_col and 'clinical_df' in globals() and age_col in getattr(clinical_df, 'columns', []):
+    miss_rate = clinical_df[age_col].isna().mean()
+    if miss_rate > 0.5:
+        age_col = None
+
+# Select gender column
+if _candidate_gender_cols:
+    if gender_preview_dict:
+        best_col = None
+        best_score = -1
+        for col, vals in gender_preview_dict.items():
+            converted = [tcga_convert_gender(v) for v in vals]
+            valid = [v for v in converted if v in (0, 1)]
+            score = len(valid)
+            if score > best_score:
+                best_col = col
+                best_score = score
+        gender_col = best_col if best_col is not None and best_score > 0 else None
+    else:
+        # Fallback by common TCGA naming
+        for preferred in ['gender', 'sex']:
+            if preferred in _candidate_gender_cols:
+                gender_col = preferred
+                break
+
+# Missingness check for gender if possible
+if gender_col and 'clinical_df' in globals() and gender_col in getattr(clinical_df, 'columns', []):
+    miss_rate = clinical_df[gender_col].isna().mean()
+    if miss_rate > 0.5:
+        gender_col = None
+
+# Explicitly print chosen columns and preview values if available
+print("Chosen age_col:", age_col)
+if age_col:
+    if 'clinical_df' in globals() and age_col in getattr(clinical_df, 'columns', []):
+        print("First 5 values of age_col:", [str(x) for x in clinical_df[age_col].head(5).tolist()])
+    elif age_preview_dict and age_col in age_preview_dict:
+        print("Preview values of age_col:", age_preview_dict[age_col])
+    else:
+        print("No preview available for age_col.")
+
+print("Chosen gender_col:", gender_col)
+if gender_col:
+    if 'clinical_df' in globals() and gender_col in getattr(clinical_df, 'columns', []):
+        print("First 5 values of gender_col:", [str(x) for x in clinical_df[gender_col].head(5).tolist()])
+    elif gender_preview_dict and gender_col in gender_preview_dict:
+        print("Preview values of gender_col:", gender_preview_dict[gender_col])
+    else:
+        print("No preview available for gender_col.")
+
+# Step 4: Feature Engineering and Validation
+import os
+import pandas as pd
+import numpy as np
+
+# 1) Extract and standardize clinical features (trait, Age, Gender)
+selected_clinical_df = tcga_select_clinical_features(
+    clinical_df=clinical_df,
+    trait=trait,
+    age_col=age_col,
+    gender_col=gender_col
+)
+
+# 2) Normalize gene symbols and save normalized gene expression
+# Ensure numeric dtype for gene expression
+gene_expr_df = genetic_df.apply(pd.to_numeric, errors='coerce')
+
+# Normalize symbols (drop unrecognized, aggregate synonyms)
+gene_expr_df = normalize_gene_symbols_in_index(gene_expr_df)
+
+# Save normalized gene expression (genes x samples)
+os.makedirs(os.path.dirname(out_gene_data_file), exist_ok=True)
+gene_expr_df.to_csv(out_gene_data_file)
+
+# 3) Link clinical and genetic data on sample IDs
+common_samples = selected_clinical_df.index.intersection(gene_expr_df.columns)
+clinical_sub = selected_clinical_df.loc[common_samples]
+gene_sub = gene_expr_df[common_samples].T  # samples x genes
+
+linked_data = pd.concat([clinical_sub, gene_sub], axis=1)
+
+# 4) Handle missing values systematically
+processed_df = handle_missing_values(linked_data, trait_col=trait)
+
+# 5) Determine bias and remove biased demographic features (trait bias used for usability)
+is_biased_flag, processed_df = judge_and_remove_biased_features(processed_df, trait)
+
+# 6) Final validation and save cohort info
+covariate_cols = [trait, 'Age', 'Gender']
+gene_cols_processed = [c for c in processed_df.columns if c not in covariate_cols]
+# Force pure Python bools to avoid JSON serialization issues
+is_gene_available = bool(len(gene_cols_processed) > 0)
+is_trait_available = bool((trait in processed_df.columns) and bool(processed_df[trait].notna().any()))
+is_biased_flag = bool(is_biased_flag)
+
+# Prepare notes
+label_counts = processed_df[trait].value_counts().to_dict() if trait in processed_df.columns else {}
+n_tumor = int(label_counts.get(1, 0))
+n_normal = int(label_counts.get(0, 0))
+note = (
+    f"INFO: Linked samples={len(processed_df)}; genes={len(gene_cols_processed)}; "
+    f"Tumor={n_tumor}; Normal={n_normal}; "
+    f"Age_included={bool('Age' in processed_df.columns)}; Gender_included={bool('Gender' in processed_df.columns)}"
+)
+
+is_usable = validate_and_save_cohort_info(
+    is_final=True,
+    cohort="TCGA",
+    info_path=json_path,
+    is_gene_available=is_gene_available,
+    is_trait_available=is_trait_available,
+    is_biased=is_biased_flag,
+    df=processed_df,
+    note=note
+)
+
+# 7) Save linked data only if usable
+if is_usable:
+    os.makedirs(os.path.dirname(out_data_file), exist_ok=True)
+    processed_df.to_csv(out_data_file)

output/preprocess/Thyroid_Cancer/cohort_info.json

@@ -1,102 +1 @@
-{
-  "GSE82208": {
-    "is_usable": true,
-    "is_gene_available": true,
-    "is_trait_available": true,
-    "is_available": true,
-    "is_biased": false,
-    "has_age": true,
-    "has_gender": true,
-    "sample_size": 52
-  },
-  "GSE80022": {
-    "is_usable": false,
-    "is_gene_available": false,
-    "is_trait_available": false,
-    "is_available": false,
-    "is_biased": null,
-    "has_age": null,
-    "has_gender": null,
-    "sample_size": null
-  },
-  "GSE76039": {
-    "is_usable": true,
-    "is_gene_available": true,
-    "is_trait_available": true,
-    "is_available": true,
-    "is_biased": false,
-    "has_age": false,
-    "has_gender": true,
-    "sample_size": 37
-  },
-  "GSE58689": {
-    "is_usable": true,
-    "is_gene_available": true,
-    "is_trait_available": true,
-    "is_available": true,
-    "is_biased": false,
-    "has_age": false,
-    "has_gender": false,
-    "sample_size": 45
-  },
-  "GSE151181": {
-    "is_usable": false,
-    "is_gene_available": false,
-    "is_trait_available": true,
-    "is_available": false,
-    "is_biased": null,
-    "has_age": null,
-    "has_gender": null,
-    "sample_size": null
-  },
-  "GSE151179": {
-    "is_usable": false,
-    "is_gene_available": true,
-    "is_trait_available": true,
-    "is_available": true,
-    "is_biased": true,
-    "has_age": false,
-    "has_gender": false,
-    "sample_size": 13
-  },
-  "GSE138198": {
-    "is_usable": false,
-    "is_gene_available": false,
-    "is_trait_available": false,
-    "is_available": false,
-    "is_biased": null,
-    "has_age": null,
-    "has_gender": null,
-    "sample_size": null
-  },
-  "GSE107754": {
-    "is_usable": false,
-    "is_gene_available": true,
-    "is_trait_available": true,
-    "is_available": true,
-    "is_biased": true,
-    "has_age": false,
-    "has_gender": true,
-    "sample_size": 84
-  },
-  "GSE104005": {
-    "is_usable": true,
-    "is_gene_available": true,
-    "is_trait_available": true,
-    "is_available": true,
-    "is_biased": false,
-    "has_age": true,
-    "has_gender": true,
-    "sample_size": 34
-  },
-  "TCGA": {
-    "is_usable": true,
-    "is_gene_available": true,
-    "is_trait_available": true,
-    "is_available": true,
-    "is_biased": false,
-    "has_age": true,
-    "has_gender": true,
-    "sample_size": 572
-  }
-}
+{"GSE82208": {"is_usable": true, "is_gene_available": true, "is_trait_available": true, "is_available": true, "is_biased": false, "has_age": true, "has_gender": true, "sample_size": 52, "note": "INFO: Gene symbols normalized using NCBI synonym map. INFO: Samples before/after missing-value handling: 52/52. INFO: Gene features before/after filtering: 19845/19845. INFO: Missing Age before handling: 2. INFO: Missing Gender before handling: 2."}, "GSE80022": {"is_usable": true, "is_gene_available": true, "is_trait_available": true, "is_available": true, "is_biased": false, "has_age": false, "has_gender": false, "sample_size": 73, "note": "INFO: Xenograft mouse model (GOT1 small intestine NET vs GOT2 medullary thyroid carcinoma); mouse hosts (Balb/c nude), no human age/gender covariates available."}, "GSE76039": {"is_usable": false, "is_gene_available": true, "is_trait_available": false, "is_available": false, "is_biased": null, "has_age": null, "has_gender": null, "sample_size": null, "note": null}, "GSE58689": {"is_usable": true, "is_gene_available": true, "is_trait_available": true, "is_available": true, "is_biased": false, "has_age": true, "has_gender": true, "sample_size": 45, "note": "INFO: Affymetrix probe IDs mapped via 'Gene Symbol' from platform SOFT; many samples lack explicit age/gender; missing values handled per protocol."}, "GSE151181": {"is_usable": false, "is_gene_available": false, "is_trait_available": false, "is_available": false, "is_biased": null, "has_age": null, "has_gender": null, "sample_size": null, "note": "INFO: Trait derived from 'tissue type' (non-neoplastic thyroid = 0, tumor/metastasis = 1)."}, "GSE151179": {"is_usable": true, "is_gene_available": true, "is_trait_available": true, "is_available": true, "is_biased": false, "has_age": false, "has_gender": false, "sample_size": 52, "note": "INFO: Age and Gender unavailable in sample characteristics; trait derived from 'tissue type'. Thermo Fisher Human Clariom S platform used; probe-to-gene mapping and symbol normalization applied."}, "GSE138198": {"is_usable": true, "is_gene_available": true, "is_trait_available": true, "is_available": true, "is_biased": false, "has_age": false, "has_gender": true, "sample_size": 23, "note": "INFO: Affymetrix Human Gene 1.0 ST platform; age not available; gender partially missing; probe-to-gene mapping via SOFT gene_assignment with equal-split aggregation."}, "GSE107754": {"is_usable": false, "is_gene_available": true, "is_trait_available": true, "is_available": true, "is_biased": true, "has_age": false, "has_gender": true, "sample_size": 71, "note": "INFO: Trait inferred from tissue; mixed metastatic tumor cohort; no age feature; gender included."}, "GSE104006": {"is_usable": false, "is_gene_available": false, "is_trait_available": true, "is_available": false, "is_biased": null, "has_age": null, "has_gender": null, "sample_size": null, "note": null}, "GSE104005": {"is_usable": true, "is_gene_available": true, "is_trait_available": true, "is_available": true, "is_biased": false, "has_age": true, "has_gender": true, "sample_size": 34, "note": "INFO: Illumina HumanHT-12 probe data mapped to gene symbols; trait from 'disease' field; miRNA platform present but gene chip used."}, "TCGA": {"is_usable": true, "is_gene_available": true, "is_trait_available": true, "is_available": true, "is_biased": false, "has_age": true, "has_gender": true, "sample_size": 572, "note": "INFO: Linked samples=572; genes=19848; Tumor=513; Normal=59; Age_included=True; Gender_included=True"}}

output/preprocess/Thyroid_Cancer/gene_data/GSE151181.csv

@@ -0,0 +1 @@
+Gene,GSM4567912,GSM4567913,GSM4567914,GSM4567915,GSM4567916,GSM4567917,GSM4567918,GSM4567919,GSM4567920,GSM4567921,GSM4567922,GSM4567923,GSM4567924,GSM4567925,GSM4567926,GSM4567927,GSM4567928,GSM4567929,GSM4567930,GSM4567931,GSM4567932,GSM4567933,GSM4567934,GSM4567935,GSM4567936,GSM4567937,GSM4567938,GSM4567939,GSM4567940,GSM4567941,GSM4567942,GSM4567943,GSM4567944,GSM4567945,GSM4567946,GSM4567947,GSM4567948,GSM4567949,GSM4567950,GSM4567951,GSM4567952,GSM4567953,GSM4567954,GSM4567955,GSM4567956,GSM4567957,GSM4567958,GSM4567959,GSM4567960,GSM4567961,GSM4567962,GSM4567963

output/preprocess/Type_1_Diabetes/clinical_data/GSE123086.csv

@@ -0,0 +1,4 @@
+,GSM3494884,GSM3494885,GSM3494886,GSM3494887,GSM3494888,GSM3494889,GSM3494890,GSM3494891,GSM3494892,GSM3494893,GSM3494894,GSM3494895,GSM3494896,GSM3494897,GSM3494898,GSM3494899,GSM3494900,GSM3494901,GSM3494902,GSM3494903,GSM3494904,GSM3494905,GSM3494906,GSM3494907,GSM3494908,GSM3494909,GSM3494910,GSM3494911,GSM3494912,GSM3494913,GSM3494914,GSM3494915,GSM3494916,GSM3494917,GSM3494918,GSM3494919,GSM3494920,GSM3494921,GSM3494922,GSM3494923,GSM3494924,GSM3494925,GSM3494926,GSM3494927,GSM3494928,GSM3494929,GSM3494930,GSM3494931,GSM3494932,GSM3494933,GSM3494934,GSM3494935,GSM3494936,GSM3494937,GSM3494938,GSM3494939,GSM3494940,GSM3494941,GSM3494942,GSM3494943,GSM3494944,GSM3494945,GSM3494946,GSM3494947,GSM3494948,GSM3494949,GSM3494950,GSM3494951,GSM3494952,GSM3494953,GSM3494954,GSM3494955,GSM3494956,GSM3494957,GSM3494958,GSM3494959,GSM3494960,GSM3494961,GSM3494962,GSM3494963,GSM3494964,GSM3494965,GSM3494966,GSM3494967,GSM3494968,GSM3494969,GSM3494970,GSM3494971,GSM3494972,GSM3494973,GSM3494974,GSM3494975,GSM3494976,GSM3494977,GSM3494978,GSM3494979,GSM3494980,GSM3494981,GSM3494982,GSM3494983,GSM3494984,GSM3494985,GSM3494986,GSM3494987,GSM3494988,GSM3494989,GSM3494990,GSM3494991,GSM3494992,GSM3494993,GSM3494994,GSM3494995,GSM3494996,GSM3494997,GSM3494998,GSM3494999,GSM3495000,GSM3495001,GSM3495002,GSM3495003,GSM3495004,GSM3495005,GSM3495006,GSM3495007,GSM3495008,GSM3495009,GSM3495010,GSM3495011,GSM3495012,GSM3495013,GSM3495014,GSM3495015,GSM3495016,GSM3495017,GSM3495018,GSM3495019,GSM3495020,GSM3495021,GSM3495022,GSM3495023,GSM3495024,GSM3495025,GSM3495026,GSM3495027,GSM3495028,GSM3495029,GSM3495030,GSM3495031,GSM3495032,GSM3495033,GSM3495034,GSM3495035,GSM3495036,GSM3495037,GSM3495038,GSM3495039,GSM3495040,GSM3495041,GSM3495042,GSM3495043,GSM3495044,GSM3495045,GSM3495046,GSM3495047,GSM3495048,GSM3495049
+Type_1_Diabetes,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,,,,,,,,,,,,,,,,
+Age,56.0,,20.0,51.0,37.0,61.0,,31.0,56.0,41.0,61.0,,80.0,53.0,61.0,73.0,60.0,76.0,77.0,74.0,69.0,77.0,81.0,70.0,82.0,69.0,82.0,67.0,67.0,78.0,67.0,74.0,,51.0,72.0,66.0,80.0,36.0,67.0,31.0,31.0,45.0,56.0,65.0,53.0,48.0,50.0,76.0,,24.0,42.0,76.0,22.0,,23.0,34.0,43.0,47.0,24.0,55.0,48.0,58.0,30.0,28.0,41.0,63.0,55.0,55.0,67.0,47.0,46.0,49.0,23.0,68.0,39.0,24.0,36.0,58.0,38.0,27.0,67.0,61.0,69.0,63.0,60.0,17.0,10.0,9.0,13.0,10.0,13.0,15.0,12.0,13.0,81.0,94.0,51.0,40.0,,97.0,23.0,93.0,58.0,28.0,54.0,15.0,8.0,11.0,12.0,8.0,14.0,8.0,10.0,14.0,13.0,40.0,52.0,42.0,29.0,43.0,41.0,54.0,42.0,49.0,45.0,56.0,64.0,71.0,48.0,20.0,53.0,32.0,26.0,28.0,47.0,24.0,48.0,,19.0,41.0,38.0,,15.0,12.0,13.0,,11.0,,16.0,11.0,,35.0,26.0,39.0,46.0,42.0,20.0,69.0,69.0,47.0,47.0,56.0,54.0,53.0,50.0,22.0
+Gender,1.0,,0.0,0.0,1.0,1.0,,1.0,0.0,0.0,0.0,,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,1.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,1.0,,1.0,1.0,1.0,0.0,1.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,1.0,,0.0,0.0,1.0,1.0,,0.0,1.0,1.0,1.0,0.0,1.0,1.0,1.0,0.0,1.0,0.0,1.0,0.0,0.0,1.0,0.0,0.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0,0.0,0.0,1.0,1.0,1.0,0.0,1.0,0.0,1.0,0.0,1.0,0.0,1.0,1.0,1.0,0.0,0.0,1.0,1.0,,1.0,1.0,0.0,1.0,0.0,1.0,1.0,1.0,1.0,1.0,0.0,1.0,0.0,1.0,1.0,1.0,0.0,0.0,0.0,0.0,0.0,0.0,1.0,1.0,1.0,0.0,1.0,0.0,0.0,0.0,1.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,,0.0,0.0,0.0,,0.0,1.0,0.0,,1.0,,1.0,1.0,,0.0,0.0,0.0,0.0,0.0,1.0,1.0,0.0,1.0,1.0,0.0,0.0,0.0,0.0,1.0

output/preprocess/Type_1_Diabetes/clinical_data/GSE123088.csv

@@ -1,2 +1,4 @@
 ,GSM3494884,GSM3494885,GSM3494886,GSM3494887,GSM3494888,GSM3494889,GSM3494890,GSM3494891,GSM3494892,GSM3494893,GSM3494894,GSM3494895,GSM3494896,GSM3494897,GSM3494898,GSM3494899,GSM3494900,GSM3494901,GSM3494902,GSM3494903,GSM3494904,GSM3494905,GSM3494906,GSM3494907,GSM3494908,GSM3494909,GSM3494910,GSM3494911,GSM3494912,GSM3494913,GSM3494914,GSM3494915,GSM3494916,GSM3494917,GSM3494918,GSM3494919,GSM3494920,GSM3494921,GSM3494922,GSM3494923,GSM3494924,GSM3494925,GSM3494926,GSM3494927,GSM3494928,GSM3494929,GSM3494930,GSM3494931,GSM3494932,GSM3494933,GSM3494934,GSM3494935,GSM3494936,GSM3494937,GSM3494938,GSM3494939,GSM3494940,GSM3494941,GSM3494942,GSM3494943,GSM3494944,GSM3494945,GSM3494946,GSM3494947,GSM3494948,GSM3494949,GSM3494950,GSM3494951,GSM3494952,GSM3494953,GSM3494954,GSM3494955,GSM3494956,GSM3494957,GSM3494958,GSM3494959,GSM3494960,GSM3494961,GSM3494962,GSM3494963,GSM3494964,GSM3494965,GSM3494966,GSM3494967,GSM3494968,GSM3494969,GSM3494970,GSM3494971,GSM3494972,GSM3494973,GSM3494974,GSM3494975,GSM3494976,GSM3494977,GSM3494978,GSM3494979,GSM3494980,GSM3494981,GSM3494982,GSM3494983,GSM3494984,GSM3494985,GSM3494986,GSM3494987,GSM3494988,GSM3494989,GSM3494990,GSM3494991,GSM3494992,GSM3494993,GSM3494994,GSM3494995,GSM3494996,GSM3494997,GSM3494998,GSM3494999,GSM3495000,GSM3495001,GSM3495002,GSM3495003,GSM3495004,GSM3495005,GSM3495006,GSM3495007,GSM3495008,GSM3495009,GSM3495010,GSM3495011,GSM3495012,GSM3495013,GSM3495014,GSM3495015,GSM3495016,GSM3495017,GSM3495018,GSM3495019,GSM3495020,GSM3495021,GSM3495022,GSM3495023,GSM3495024,GSM3495025,GSM3495026,GSM3495027,GSM3495028,GSM3495029,GSM3495030,GSM3495031,GSM3495032,GSM3495033,GSM3495034,GSM3495035,GSM3495036,GSM3495037,GSM3495038,GSM3495039,GSM3495040,GSM3495041,GSM3495042,GSM3495043,GSM3495044,GSM3495045,GSM3495046,GSM3495047,GSM3495048,GSM3495049,GSM3495050,GSM3495051,GSM3495052,GSM3495053,GSM3495054,GSM3495055,GSM3495056,GSM3495057,GSM3495058,GSM3495059,GSM3495060,GSM3495061,GSM3495062,GSM3495063,GSM3495064,GSM3495065,GSM3495066,GSM3495067,GSM3495068,GSM3495069,GSM3495070,GSM3495071,GSM3495072,GSM3495073,GSM3495074,GSM3495075,GSM3495076,GSM3495077,GSM3495078,GSM3495079,GSM3495080,GSM3495081,GSM3495082,GSM3495083,GSM3495084,GSM3495085,GSM3495086,GSM3495087
-Type_1_Diabetes,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
+Type_1_Diabetes,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0
+Age,56.0,,20.0,51.0,37.0,61.0,,31.0,56.0,41.0,61.0,,80.0,53.0,61.0,73.0,60.0,76.0,77.0,74.0,69.0,77.0,81.0,70.0,82.0,69.0,82.0,67.0,67.0,78.0,67.0,74.0,,51.0,72.0,66.0,80.0,36.0,67.0,31.0,31.0,45.0,56.0,65.0,53.0,48.0,50.0,76.0,,24.0,42.0,76.0,22.0,,23.0,34.0,43.0,47.0,24.0,55.0,48.0,58.0,30.0,28.0,41.0,63.0,55.0,55.0,67.0,47.0,46.0,49.0,23.0,68.0,39.0,24.0,36.0,58.0,38.0,27.0,67.0,61.0,69.0,63.0,60.0,17.0,10.0,9.0,13.0,10.0,13.0,15.0,12.0,13.0,81.0,94.0,51.0,40.0,,97.0,23.0,93.0,58.0,28.0,54.0,15.0,8.0,11.0,12.0,8.0,14.0,8.0,10.0,14.0,13.0,40.0,52.0,42.0,29.0,43.0,41.0,54.0,42.0,49.0,45.0,56.0,64.0,71.0,48.0,20.0,53.0,32.0,26.0,28.0,47.0,24.0,48.0,,19.0,41.0,38.0,,15.0,12.0,13.0,,11.0,,16.0,11.0,,35.0,26.0,39.0,46.0,42.0,20.0,69.0,69.0,47.0,47.0,56.0,54.0,53.0,50.0,22.0,62.0,74.0,57.0,47.0,70.0,50.0,52.0,43.0,57.0,53.0,70.0,41.0,61.0,39.0,58.0,55.0,63.0,60.0,43.0,68.0,67.0,50.0,67.0,51.0,59.0,44.0,35.0,83.0,78.0,88.0,41.0,60.0,72.0,53.0,73.0,56.0,38.0,53.0
+Gender,1.0,,0.0,0.0,1.0,1.0,,1.0,0.0,0.0,0.0,,1.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,1.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,1.0,,1.0,1.0,1.0,0.0,1.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,1.0,,0.0,0.0,1.0,1.0,,0.0,1.0,1.0,1.0,0.0,1.0,1.0,1.0,0.0,1.0,0.0,1.0,0.0,0.0,1.0,0.0,0.0,1.0,1.0,1.0,1.0,1.0,0.0,0.0,0.0,0.0,0.0,1.0,1.0,1.0,0.0,1.0,0.0,1.0,0.0,1.0,0.0,1.0,1.0,1.0,0.0,0.0,1.0,1.0,,1.0,1.0,0.0,1.0,0.0,1.0,1.0,1.0,1.0,1.0,0.0,1.0,0.0,1.0,1.0,1.0,0.0,0.0,0.0,0.0,0.0,0.0,1.0,1.0,1.0,0.0,1.0,0.0,0.0,0.0,1.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,,0.0,0.0,0.0,,0.0,1.0,0.0,,1.0,,1.0,1.0,,0.0,0.0,0.0,0.0,0.0,1.0,1.0,0.0,1.0,1.0,0.0,0.0,0.0,0.0,1.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0

output/preprocess/Type_1_Diabetes/clinical_data/GSE156035.csv

@@ -1,3 +1,3 @@
-GSM4720871,GSM4720872,GSM4720873,GSM4720874,GSM4720875,GSM4720876,GSM4720877,GSM4720878,GSM4720879,GSM4720880,GSM4720881,GSM4720882,GSM4720883,GSM4720884,GSM4720885,GSM4720886,GSM4720887,GSM4720888,GSM4720889,GSM4720890,GSM4720891,GSM4720892,GSM4720893,GSM4720894,GSM4720895,GSM4720896,GSM4720897,GSM4720898,GSM4720899,GSM4720900,GSM4720901,GSM4720902,GSM4720903,GSM4720904,GSM4720905,GSM4720906,GSM4720907,GSM4720908,GSM4720909,GSM4720910
-0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0
-0.0,0.0,1.0,0.0,1.0,0.0,1.0,1.0,0.0,0.0,1.0,0.0,0.0,0.0,0.0,0.0,1.0,1.0,0.0,0.0,0.0,0.0,1.0,1.0,1.0,0.0,0.0,0.0,0.0,1.0,1.0,0.0,0.0,0.0,0.0,0.0,1.0,0.0,1.0,1.0
+,GSM4720871,GSM4720872,GSM4720873,GSM4720874,GSM4720875,GSM4720876,GSM4720877,GSM4720878,GSM4720879,GSM4720880,GSM4720881,GSM4720882,GSM4720883,GSM4720884,GSM4720885,GSM4720886,GSM4720887,GSM4720888,GSM4720889,GSM4720890,GSM4720891,GSM4720892,GSM4720893,GSM4720894,GSM4720895,GSM4720896,GSM4720897,GSM4720898,GSM4720899,GSM4720900,GSM4720901,GSM4720902,GSM4720903,GSM4720904,GSM4720905,GSM4720906,GSM4720907,GSM4720908,GSM4720909,GSM4720910
+Type_1_Diabetes,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0
+Gender,0.0,0.0,1.0,0.0,1.0,0.0,1.0,1.0,0.0,0.0,1.0,0.0,0.0,0.0,0.0,0.0,1.0,1.0,0.0,0.0,0.0,0.0,1.0,1.0,1.0,0.0,0.0,0.0,0.0,1.0,1.0,0.0,0.0,0.0,0.0,0.0,1.0,0.0,1.0,1.0

output/preprocess/Type_1_Diabetes/clinical_data/GSE193273.csv

@@ -1,2 +1,2 @@
-,0,1
-Type_1_Diabetes,0.0,1.0
+,GSM5781088,GSM5781089,GSM5781090,GSM5781091,GSM5781092,GSM5781093,GSM5781095,GSM5781096,GSM5781097,GSM5781098,GSM5781099,GSM5781100,GSM5781101,GSM5781102,GSM5781104,GSM5781105,GSM5781106,GSM5781107,GSM5781108,GSM5781109,GSM5781110,GSM5781111,GSM5781113,GSM5781114,GSM5781115,GSM5781116,GSM5781117,GSM5781118,GSM5781119,GSM5781120,GSM5781121,GSM5781123,GSM5781124,GSM5781125,GSM5781126,GSM5781127,GSM5781128,GSM5781129,GSM5781131,GSM5781132
+Type_1_Diabetes,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0,1.0

output/preprocess/Type_1_Diabetes/code/GSE123086.py

@@ -0,0 +1,322 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Type_1_Diabetes"
+cohort = "GSE123086"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Type_1_Diabetes"
+in_cohort_dir = "../DATA/GEO/Type_1_Diabetes/GSE123086"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Type_1_Diabetes/GSE123086.csv"
+out_gene_data_file = "./output/z6/preprocess/Type_1_Diabetes/gene_data/GSE123086.csv"
+out_clinical_data_file = "./output/z6/preprocess/Type_1_Diabetes/clinical_data/GSE123086.csv"
+json_path = "./output/z6/preprocess/Type_1_Diabetes/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
+import re
+import pandas as pd
+
+# 1) Gene expression data availability
+is_gene_available = True  # Agilent one-color gene expression microarray per Series_overall_design
+
+# 2) Variable availability and converters based on Sample Characteristics Dictionary
+trait_row = 1   # 'primary diagnosis: ...' includes TYPE_1_DIABETES and HEALTHY_CONTROL
+age_row = 3     # contains many 'age: ...' entries (more coverage than row 4)
+gender_row = 2  # contains 'Sex: Female'/'Sex: Male' entries
+
+def _split_header_value(x: str):
+    if x is None:
+        return None, None
+    parts = str(x).split(":", 1)
+    if len(parts) == 2:
+        header = parts[0].strip().lower()
+        value = parts[1].strip()
+        return header, value
+    # If no colon, treat entire string as value with unknown header
+    return None, str(x).strip()
+
+def convert_trait(x):
+    header, val = _split_header_value(x)
+    if val is None:
+        return None
+    # Only use 'primary diagnosis' field for trait conversion
+    if header not in {"primary diagnosis", "primary_diagnosis", "primarydiagnosis"}:
+        return None
+    v = val.upper().replace("-", "_").replace(" ", "_")
+    # Map T1D cases to 1
+    if ("TYPE_1_DIABETES" in v) or ("TYPE_I_DIABETES" in v):
+        return 1
+    # Map healthy controls to 0
+    if ("HEALTHY" in v) or ("CONTROL" in v) or ("HEALTHY_CONTROL" in v):
+        return 0
+    # Other diseases -> not applicable for this trait
+    return None
+
+def convert_age(x):
+    header, val = _split_header_value(x)
+    if val is None:
+        return None
+    if header != "age":
+        return None
+    # Extract first numeric token
+    m = re.search(r"[-+]?\d+(\.\d+)?", val)
+    if not m:
+        return None
+    try:
+        age = float(m.group())
+    except Exception:
+        return None
+    if 0 <= age <= 120:
+        return age
+    return None
+
+def convert_gender(x):
+    header, val = _split_header_value(x)
+    if val is None:
+        return None
+    if header not in {"sex", "gender"}:
+        return None
+    v = val.strip().lower()
+    if v in {"male", "m"}:
+        return 1
+    if v in {"female", "f"}:
+        return 0
+    return None
+
+# 3) Initial filtering 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
+)
+
+# 4) Clinical feature extraction, preview, and save
+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)
+
+    os.makedirs(os.path.dirname(out_clinical_data_file), exist_ok=True)
+    selected_clinical_df.to_csv(out_clinical_data_file)
+
+# 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
+# Based on the observed identifiers (numeric IDs like '1','2','3',...), these are not human gene symbols.
+requires_gene_mapping = True
+print(f"requires_gene_mapping = {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
+import re
+import pandas as pd
+
+# Inspect available columns to choose ID and symbol columns robustly
+ann_cols = list(gene_annotation.columns)
+
+# 1) Decide identifier and gene symbol columns
+expr_ids = set(gene_data.index.astype(str))
+
+# Choose probe/ID column: prefer 'ID' if present, otherwise choose max overlap
+if 'ID' in gene_annotation.columns:
+    id_col = 'ID'
+else:
+    overlap_scores = {}
+    for col in ann_cols:
+        vals = gene_annotation[col].astype(str)
+        overlap_scores[col] = (vals.head(2000).isin(expr_ids)).mean()
+    id_col = max(overlap_scores, key=overlap_scores.get)
+
+# Try to find a gene symbol-like column using flexible matching
+symbol_candidates = [
+    'GENE_SYMBOL', 'GENE SYMBOL', 'GENE_SYMBOLS', 'GENE SYMBOLS',
+    'GENE_SYMBOL_CH1', 'GENE SYMBOL CH1', 'SYMBOL', 'GENE', 'GENE_NAME', 'GENE NAME',
+    'GENE_TITLE', 'GENE TITLE', 'HGNC_SYMBOL', 'HGNC SYMBOL', 'GENES'
+]
+def normalize_colname(c):
+    return re.sub(r'[^A-Z0-9]', '', str(c).upper())
+normalized_map = {normalize_colname(c): c for c in ann_cols}
+symbol_col = None
+for cand in symbol_candidates:
+    norm_cand = normalize_colname(cand)
+    if norm_cand in normalized_map:
+        symbol_col = normalized_map[norm_cand]
+        break
+
+print(f"[INFO] Using ID column: {id_col}")
+print(f"[INFO] Gene symbol column: {symbol_col if symbol_col is not None else 'None (will fallback to ENTREZ_GENE_ID)'}")
+
+# 2) Build mapping dataframe and 3) Apply mapping to convert to gene-level data
+if symbol_col is not None:
+    # Use provided helper for mapping extraction; apply_gene_mapping will parse symbols sensibly
+    mapping_df = get_gene_mapping(gene_annotation, prob_col=id_col, gene_col=symbol_col)
+    gene_data = apply_gene_mapping(gene_data, mapping_df)
+else:
+    # Fallback: use ENTREZ_GENE_ID as gene identifier (do not blindly split; only split on clear multi-maps)
+    if 'ENTREZ_GENE_ID' not in gene_annotation.columns:
+        # If even ENTREZ is not available, map 1:1 to probe IDs to avoid explosion
+        print("[WARNING] ENTREZ_GENE_ID not found. Keeping probe-level data (no mapping applied).")
+        # Keep gene_data as is (probe-level). This is a conservative fallback.
+    else:
+        temp_map = gene_annotation.loc[:, [id_col, 'ENTREZ_GENE_ID']].copy()
+        temp_map.columns = ['ID', 'GeneRaw']
+        temp_map = temp_map.dropna()
+        temp_map['ID'] = temp_map['ID'].astype(str).str.strip()
+        temp_map['GeneRaw'] = temp_map['GeneRaw'].astype(str).str.strip()
+
+        # Keep only IDs present in expression data
+        temp_map = temp_map[temp_map['ID'].isin(gene_data.index)]
+
+        # Split only if clear multi-mapping separators exist; keep only numeric Entrez tokens
+        sep_pattern = re.compile(r'///|[,;|]')
+        def to_gene_list(s: str):
+            if pd.isna(s):
+                return []
+            s = str(s).strip()
+            if not s:
+                return []
+            if sep_pattern.search(s):
+                parts = re.split(r'\s*(?:///|[,;|])\s*', s)
+            else:
+                parts = [s]
+            # Keep only strictly numeric Entrez IDs and remove zeros or empty tokens
+            parts = [p for p in parts if re.fullmatch(r'\d+', p) and p != '0']
+            # Deduplicate while preserving order
+            return list(dict.fromkeys(parts))
+
+        temp_map['GeneList'] = temp_map['GeneRaw'].map(to_gene_list)
+        # Drop mappings with no valid Entrez ID
+        temp_map = temp_map[temp_map['GeneList'].map(len) > 0].copy()
+
+        # Compute number of genes per probe
+        temp_map['num_genes'] = temp_map['GeneList'].map(len)
+
+        # Explode to one row per probe-gene pair and deduplicate
+        mapping_df = temp_map[['ID', 'GeneList', 'num_genes']].explode('GeneList')
+        mapping_df = mapping_df.rename(columns={'GeneList': 'Gene'})
+        mapping_df = mapping_df.drop_duplicates(subset=['ID', 'Gene'])
+
+        # Join with expression and distribute equally
+        mapping_df = mapping_df.set_index('ID')
+        merged = mapping_df.join(gene_data, how='inner')
+        expr_cols = [c for c in merged.columns if c not in ['Gene', 'num_genes']]
+
+        # Safety: avoid division by zero
+        merged['num_genes'] = merged['num_genes'].replace(0, 1)
+        merged[expr_cols] = merged[expr_cols].div(merged['num_genes'], axis=0)
+
+        # Aggregate to gene (Entrez) level
+        gene_data = merged.groupby('Gene')[expr_cols].sum()
+
+# Sanity checks
+print(f"[INFO] Gene-level data shape: {gene_data.shape}")
+if isinstance(gene_data.index, pd.Index):
+    print(f"[INFO] Unique gene identifiers: {gene_data.index.nunique()}")
+
+# Step 7: Data Normalization and Linking
+import os
+import pandas as pd
+
+# 1. Normalize gene symbols if appropriate; otherwise keep Entrez-level data and leave a note.
+note = ""
+# Detect if index looks like Entrez IDs (all numeric)
+idx_as_str = pd.Index(gene_data.index.astype(str))
+is_entrez_index = idx_as_str.str.fullmatch(r'\d+').all()
+
+if is_entrez_index:
+    # Skip normalization due to lack of gene symbols in annotation; keep Entrez-level data
+    normalized_gene_data = gene_data
+    note = "WARNING: Skipped gene-symbol normalization because gene index consists of Entrez IDs and no gene symbol annotations were available."
+else:
+    # Attempt normalization; if it collapses too much, fall back to original
+    temp_normalized = normalize_gene_symbols_in_index(gene_data.copy())
+    if temp_normalized.shape[0] == 0:
+        normalized_gene_data = gene_data
+        note = "WARNING: Gene-symbol normalization produced an empty matrix; reverted to original gene data."
+    else:
+        normalized_gene_data = temp_normalized
+        note = "INFO: Gene symbols normalized using synonym mapping."
+
+# Save gene (possibly normalized) data
+os.makedirs(os.path.dirname(out_gene_data_file), exist_ok=True)
+normalized_gene_data.to_csv(out_gene_data_file)
+
+# 2. Link the clinical and genetic data
+# Fix variable name to use the existing clinical dataframe from Step 2
+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. Determine bias and remove biased demographics
+is_trait_biased, unbiased_linked_data = judge_and_remove_biased_features(linked_data, trait)
+
+# 5. Final validation and save cohort info
+# Determine availability flags based on data presence
+is_gene_available_flag = (normalized_gene_data.shape[0] > 0 and normalized_gene_data.shape[1] > 0)
+# Trait data was extracted earlier; let final validation handle abnormalities if any
+is_trait_available_flag = True
+
+is_usable = validate_and_save_cohort_info(
+    is_final=True,
+    cohort=cohort,
+    info_path=json_path,
+    is_gene_available=is_gene_available_flag,
+    is_trait_available=is_trait_available_flag,
+    is_biased=is_trait_biased,
+    df=unbiased_linked_data,
+    note=note
+)
+
+# 6. Save linked data only if usable
+if is_usable:
+    os.makedirs(os.path.dirname(out_data_file), exist_ok=True)
+    unbiased_linked_data.to_csv(out_data_file)

output/preprocess/Type_1_Diabetes/code/GSE123088.py

@@ -0,0 +1,303 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Type_1_Diabetes"
+cohort = "GSE123088"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Type_1_Diabetes"
+in_cohort_dir = "../DATA/GEO/Type_1_Diabetes/GSE123088"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Type_1_Diabetes/GSE123088.csv"
+out_gene_data_file = "./output/z6/preprocess/Type_1_Diabetes/gene_data/GSE123088.csv"
+out_clinical_data_file = "./output/z6/preprocess/Type_1_Diabetes/clinical_data/GSE123088.csv"
+json_path = "./output/z6/preprocess/Type_1_Diabetes/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
+import re
+
+# 1. Gene Expression Data Availability
+is_gene_available = True  # Based on series context indicating single-cell gene expression (not miRNA/methylation)
+
+# 2. Variable Availability and Data Type Conversion
+
+# Identify rows where variables are recorded
+trait_row = 1     # 'primary diagnosis: ...'
+age_row = 3       # 'age: ...' (most ages appear here)
+gender_row = 2    # 'Sex: Female/Male'
+
+# Conversion functions
+def convert_trait(x):
+    if x is None:
+        return None
+    s = str(x).strip()
+    parts = s.split(":", 1)
+    header = parts[0].strip().lower()
+    value = parts[1].strip() if len(parts) > 1 else s.strip()
+    v = value.lower()
+
+    if 'primary' in header and 'diagnosis' in header:
+        # Map presence of Type 1 Diabetes to 1, others (including controls and other diseases) to 0
+        v_std = re.sub(r'[^a-z0-9]+', '_', v)
+        if ('diab' in v and ('type_1' in v_std or 'type1' in v_std or 't1d' in v_std)):
+            return 1
+        else:
+            return 0
+    return None
+
+def convert_age(x):
+    if x is None:
+        return None
+    s = str(x).strip()
+    parts = s.split(":", 1)
+    header = parts[0].strip().lower()
+    value = parts[1].strip() if len(parts) > 1 else s.strip()
+    if 'age' in header:
+        m = re.search(r'[-+]?\d*\.?\d+', value)
+        if m:
+            try:
+                val = float(m.group())
+                return val
+            except:
+                return None
+    return None
+
+def convert_gender(x):
+    if x is None:
+        return None
+    s = str(x).strip()
+    parts = s.split(":", 1)
+    header = parts[0].strip().lower()
+    value = parts[1].strip() if len(parts) > 1 else s.strip()
+    if ('sex' in header) or ('gender' in header):
+        v = value.lower()
+        if 'female' in v:
+            return 0
+        if 'male' in v:
+            return 1
+        return None
+    return None
+
+# 3. Save Metadata (initial filtering)
+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
+)
+
+# 4. Clinical Feature Extraction (only if trait 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)
+
+    os.makedirs(os.path.dirname(out_clinical_data_file), exist_ok=True)
+    selected_clinical_df.to_csv(out_clinical_data_file)
+
+# 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
+requires_gene_mapping = True
+print(f"requires_gene_mapping = {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
+import re
+import pandas as pd
+
+# Recreate probe-level expression to avoid contamination from prior attempts
+expr_df = get_genetic_data(matrix_file)
+
+# Try to find a gene symbol column in the annotation
+symbol_col = None
+colnames = list(gene_annotation.columns)
+
+# Common symbol-like names
+preferred_names = [
+    'Gene Symbol', 'GENE_SYMBOL', 'SYMBOL', 'Gene symbol', 'Symbol', 'gene_symbol', 'gene symbol'
+]
+
+# First pass: exact-ish matches by name
+for name in preferred_names:
+    if name in colnames:
+        symbol_col = name
+        break
+
+# Second pass: regex match on column names if not found
+if symbol_col is None:
+    for c in colnames:
+        if re.search(r'\b(gene\s*symbol|symbol)\b', str(c), flags=re.I):
+            symbol_col = c
+            break
+
+# Validate candidate symbol column by content (must yield plausible human gene symbols)
+def looks_like_symbol_col(series: pd.Series, sample_n: int = 200) -> bool:
+    # sample a subset to avoid heavy computation
+    sample_vals = series.astype(str).head(sample_n)
+    count_symbolish = 0
+    for v in sample_vals:
+        syms = extract_human_gene_symbols(v)
+        if len(syms) > 0:
+            count_symbolish += 1
+    # Heuristic: at least 10% rows produce plausible symbols
+    return (count_symbolish / max(1, len(sample_vals))) >= 0.1
+
+if symbol_col is not None and not looks_like_symbol_col(gene_annotation[symbol_col]):
+    symbol_col = None
+
+if symbol_col is not None:
+    # Use library helpers to map to gene symbols and aggregate
+    mapping_df = get_gene_mapping(annotation=gene_annotation, prob_col='ID', gene_col=symbol_col)
+    mapping_df = mapping_df[mapping_df['ID'].isin(expr_df.index)]
+    print(f"Using symbol column: {symbol_col}. Candidate mappings: {len(mapping_df)}")
+
+    gene_data = apply_gene_mapping(expression_df=expr_df, mapping_df=mapping_df)
+
+    # Sanity checks
+    n_genes = gene_data.shape[0]
+    if n_genes <= 0 or n_genes > 70000:
+        raise ValueError(f"Unreasonable number of genes after symbol mapping: {n_genes}")
+    print(f"Gene-level matrix (symbol) shape: {gene_data.shape}")
+    print(f"First 5 gene symbols: {list(gene_data.index[:5])}")
+
+else:
+    # Fallback: robust Entrez ID mapping
+    if 'ENTREZ_GENE_ID' not in gene_annotation.columns:
+        raise ValueError("No gene symbol column found and ENTREZ_GENE_ID column is missing; cannot build mapping.")
+
+    mapping_df = gene_annotation.loc[:, ['ID', 'ENTREZ_GENE_ID']].dropna().copy()
+    mapping_df['ID'] = mapping_df['ID'].astype(str).str.strip()
+    mapping_df = mapping_df[mapping_df['ID'].isin(expr_df.index)]
+
+    # Restrictive split: GEO uses "///" for multi-mapping; also accept semicolon/comma
+    def split_entrez_ids(x: str):
+        if not isinstance(x, str):
+            return []
+        parts = re.split(r'\s*///\s*|[;,]+', x)
+        parts = [p.strip() for p in parts if p and p.upper() not in {'NA', 'NAN', '---'}]
+        # Keep strictly numeric (Entrez IDs)
+        parts = [p for p in parts if re.fullmatch(r'\d+', p) is not None]
+        return parts
+
+    mapping_df['Gene'] = mapping_df['ENTREZ_GENE_ID'].astype(str).apply(split_entrez_ids)
+    mapping_df['num_genes'] = mapping_df['Gene'].apply(len)
+    mapping_df = mapping_df[mapping_df['num_genes'] > 0]
+    mapping_df = mapping_df.explode('Gene').dropna(subset=['Gene'])
+
+    # Make Gene labels explicit and non-numeric to avoid collision/contamination
+    mapping_df['Gene'] = mapping_df['Gene'].astype(str).map(lambda x: f"ENTREZ:{x}")
+
+    # Join and aggregate
+    mapping_df = mapping_df.set_index('ID')
+    merged = mapping_df.join(expr_df, how='inner')
+
+    expr_cols = [c for c in merged.columns if c not in ['ENTREZ_GENE_ID', 'Gene', 'num_genes']]
+    merged[expr_cols] = merged[expr_cols].div(merged['num_genes'].replace(0, 1), axis=0)
+
+    gene_data = merged.groupby('Gene', sort=False)[expr_cols].sum()
+
+    # Sanity checks
+    n_genes = gene_data.shape[0]
+    if n_genes <= 0 or n_genes > 70000:
+        raise ValueError(f"Unreasonable number of genes after Entrez mapping: {n_genes}")
+    # Ensure IDs look like "ENTREZ:####"
+    bad_ids = [g for g in gene_data.index[:1000] if re.fullmatch(r'ENTREZ:\d+', g) is None]
+    if len(bad_ids) > 0:
+        raise ValueError(f"Unexpected gene IDs detected (showing up to 5): {bad_ids[:5]}")
+
+    print(f"Gene-level matrix (Entrez) shape: {gene_data.shape}")
+    print(f"First 5 gene IDs: {list(gene_data.index[:5])}")
+
+# Step 7: Data Normalization and Linking
+import os
+import re
+
+# 1. Normalize gene symbols if applicable; skip if IDs are Entrez
+entrez_pattern = re.compile(r'^ENTREZ:\d+$')
+all_entrez = all(entrez_pattern.fullmatch(idx) is not None for idx in gene_data.index)
+
+note = ""
+if all_entrez:
+    # Skip normalization because we have Entrez-style IDs, not gene symbols
+    normalized_gene_data = gene_data.copy()
+    note = "INFO: Gene normalization skipped because gene IDs are Entrez (ENTREZ:####)."
+else:
+    normalized_gene_data = normalize_gene_symbols_in_index(gene_data)
+    note = "INFO: Gene symbols normalized using NCBI synonym mapping."
+
+# Ensure output directory exists and save 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 the clinical and genetic data
+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 demographic features are severely biased, and remove biased features
+is_trait_biased, unbiased_linked_data = judge_and_remove_biased_features(linked_data, trait)
+
+# 5. Final quality validation and save cohort information
+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,
+    note=note
+)
+
+# 6. Save linked data only if usable
+if is_usable:
+    os.makedirs(os.path.dirname(out_data_file), exist_ok=True)
+    unbiased_linked_data.to_csv(out_data_file)

output/preprocess/Type_1_Diabetes/code/GSE131528.py

@@ -0,0 +1,204 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Type_1_Diabetes"
+cohort = "GSE131528"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Type_1_Diabetes"
+in_cohort_dir = "../DATA/GEO/Type_1_Diabetes/GSE131528"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Type_1_Diabetes/GSE131528.csv"
+out_gene_data_file = "./output/z6/preprocess/Type_1_Diabetes/gene_data/GSE131528.csv"
+out_clinical_data_file = "./output/z6/preprocess/Type_1_Diabetes/clinical_data/GSE131528.csv"
+json_path = "./output/z6/preprocess/Type_1_Diabetes/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
+# Determine data availability based on provided background and characteristics
+
+# 1) Gene expression availability: Title suggests immune signature profiling; likely gene expression (not miRNA-only or methylation)
+is_gene_available = True
+
+# 2) Variable availability and conversion functions
+
+# Trait (Type_1_Diabetes): Only T1D/new onset T1D observed, no healthy/control in the sample characteristics.
+# Treat as not available for our binary case-control analysis.
+trait_row = None  # constant "case-only" cohort is not useful for association with the trait
+
+# Age: Available in key 3 (age at enrollment / age at blood draw)
+age_row = 3
+
+# Gender: Available in key 5 (sex: female/male)
+gender_row = 5
+
+# Conversion utilities
+def _after_colon(value):
+    if value is None:
+        return None
+    s = str(value)
+    if ':' in s:
+        s = s.split(':', 1)[1]
+    return s.strip()
+
+def convert_trait(value):
+    v = _after_colon(value)
+    if v is None:
+        return None
+    vl = v.lower()
+    # Positive (T1D) indicators
+    positive = ['t1d', 'type 1 diabetes', 'type i diabetes', 'new onset t1d', 'diabetic']
+    if any(tok in vl for tok in positive):
+        return 1
+    # Negative (control/healthy) indicators
+    negative = ['control', 'healthy', 'non-diabetic', 'no diabetes', 'normoglycemic', 'normoglycaemic']
+    if any(tok in vl for tok in negative):
+        return 0
+    return None
+
+def convert_age(value):
+    v = _after_colon(value)
+    if v is None:
+        return None
+    vl = v.lower()
+    # Only attempt to parse when the field is about age; otherwise return None
+    if 'age' not in vl:
+        # Some entries in the same row may be race or other descriptors
+        # If the part after colon is numeric already (rare), still try to parse
+        pass
+    # Handle NA-like values
+    if v.strip().lower() in ['na', 'n/a', '', 'nan']:
+        return None
+    # Extract numeric token (age values look like plain numbers)
+    try:
+        return float(v)
+    except Exception:
+        # Try to extract leading/trailing numeric substrings
+        import re
+        m = re.search(r'[-+]?\d*\.?\d+', v)
+        if m:
+            try:
+                return float(m.group(0))
+            except Exception:
+                return None
+        return None
+
+def convert_gender(value):
+    v = _after_colon(value)
+    if v is None:
+        return None
+    vl = v.lower()
+    # Accept various representations
+    if vl in ['f', 'female']:
+        return 0
+    if vl in ['m', 'male']:
+        return 1
+    # Some entries in the same row are unrelated (e.g., AUC); return None
+    return None
+
+# 3) Save metadata (initial filtering)
+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)
+
+# 4) Clinical feature extraction: skip because trait_row is None (trait not available for this dataset)
+# If in future trait_row becomes available, the following template can be used:
+# selected_clinical_df = geo_select_clinical_features(clinical_data, trait, trait_row,
+#                                                     convert_trait,
+#                                                     age_row=age_row,
+#                                                     convert_age=convert_age,
+#                                                     gender_row=gender_row,
+#                                                     convert_gender=convert_gender)
+# preview = preview_df(selected_clinical_df)
+# os.makedirs(os.path.dirname(out_clinical_data_file), exist_ok=True)
+# selected_clinical_df.to_csv(out_clinical_data_file)
+
+# 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
+requires_gene_mapping = True
+print(f"requires_gene_mapping = {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
+# 1-2. Decide columns and build mapping dataframe
+probe_col = 'ID'  # matches probe identifiers in gene_data index
+gene_symbol_col = 'Gene Symbol'  # contains gene symbols
+mapping_df = get_gene_mapping(gene_annotation, prob_col=probe_col, gene_col=gene_symbol_col)
+
+# 3. Apply mapping to convert probe-level data to gene-level expression
+gene_data = apply_gene_mapping(expression_df=gene_data, mapping_df=mapping_df)
+
+# Step 7: Data Normalization and Linking
+import os
+
+# 1. Normalize gene symbols and save gene-level 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)
+
+# 2-6. Proceed only if clinical features were extracted earlier; otherwise skip linking and final validation
+if 'selected_clinical_data' in globals() and selected_clinical_data is not None:
+    # 2. Link clinical and genetic data
+    linked_data = geo_link_clinical_genetic_data(selected_clinical_data, normalized_gene_data)
+
+    # 3. Handle missing values
+    linked_data = handle_missing_values(linked_data, trait)
+
+    # 4. Bias checks
+    is_trait_biased, unbiased_linked_data = judge_and_remove_biased_features(linked_data, trait)
+
+    # 5. Final validation and cohort info
+    is_usable = validate_and_save_cohort_info(
+        True, cohort, json_path, True, True, is_trait_biased, unbiased_linked_data
+    )
+
+    # 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)
+else:
+    # Trait not available; record initial metadata (if not already done) and skip linking
+    _ = validate_and_save_cohort_info(
+        is_final=False,
+        cohort=cohort,
+        info_path=json_path,
+        is_gene_available=True,
+        is_trait_available=False
+    )

output/preprocess/Type_1_Diabetes/code/GSE156035.py

@@ -0,0 +1,197 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Type_1_Diabetes"
+cohort = "GSE156035"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Type_1_Diabetes"
+in_cohort_dir = "../DATA/GEO/Type_1_Diabetes/GSE156035"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Type_1_Diabetes/GSE156035.csv"
+out_gene_data_file = "./output/z6/preprocess/Type_1_Diabetes/gene_data/GSE156035.csv"
+out_clinical_data_file = "./output/z6/preprocess/Type_1_Diabetes/clinical_data/GSE156035.csv"
+json_path = "./output/z6/preprocess/Type_1_Diabetes/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 data availability
+is_gene_available = True  # Global gene expression profiling in PBMC (not miRNA/methylation)
+
+# Step 2: Identify rows for variables based on Sample Characteristics Dictionary
+trait_row = 2    # 'diagnosis: Healthy control' vs 'diagnosis: Type 1 diabetes'
+age_row = None   # Not available in provided characteristics
+gender_row = 0   # 'gender: Female/Male'
+
+# Step 2.2: Define conversion functions
+def _after_colon(value):
+    if value is None:
+        return None
+    parts = str(value).split(":", 1)
+    v = parts[-1].strip() if len(parts) > 1 else str(value).strip()
+    return v
+
+def convert_trait(x):
+    v = _after_colon(x)
+    if v is None or v == '':
+        return None
+    vl = v.lower()
+    # Map T1D cases to 1, controls to 0
+    if any(k in vl for k in ["type 1 diabetes", "t1d", "type1", "type i diabetes", "recent-onset type 1"]):
+        return 1
+    if any(k in vl for k in ["healthy control", "control", "healthy", "non-diabetic", "islet autoantibody-negative healthy controls"]):
+        return 0
+    # Fallback heuristics
+    if any(k in vl for k in ["case", "patient", "diabetes"]):
+        return 1
+    if "control" in vl:
+        return 0
+    return None
+
+def convert_age(x):
+    v = _after_colon(x)
+    if v is None or v == '':
+        return None
+    vl = v.lower().replace("years", "").replace("year", "").replace("yrs", "").replace("yr", "").strip()
+    # Remove common non-numeric characters
+    vl = ''.join(ch for ch in vl if (ch.isdigit() or ch == '.' or ch == '-'))
+    try:
+        return float(vl)
+    except:
+        return None
+
+def convert_gender(x):
+    v = _after_colon(x)
+    if v is None or v == '':
+        return None
+    vl = v.strip().lower()
+    if vl in ["female", "f", "woman", "women"]:
+        return 0
+    if vl in ["male", "m", "man", "men"]:
+        return 1
+    return None
+
+# Step 3: Initial filtering 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 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 if age_row is not None else None,
+        gender_row=gender_row,
+        convert_gender=convert_gender
+    )
+    preview = preview_df(selected_clinical_df, n=5)
+    print(preview)
+    # Save clinical data
+    os.makedirs(os.path.dirname(out_clinical_data_file), exist_ok=True)
+    selected_clinical_df.to_csv(out_clinical_data_file)
+
+# 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
+print("requires_gene_mapping = True")
+
+# 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
+# Identify the appropriate columns in the annotation:
+# - Probe/row identifiers in expression data match the 'ID' column in annotation
+# - Gene symbols are in the 'GENE_SYMBOL' column
+probe_col = 'ID'
+gene_col = 'GENE_SYMBOL'
+
+# Build mapping dataframe
+mapping_df = get_gene_mapping(gene_annotation, prob_col=probe_col, gene_col=gene_col)
+
+# Apply mapping to convert probe-level data to gene-level data
+gene_data = apply_gene_mapping(gene_data, mapping_df)
+
+# Step 7: Data Normalization and Linking
+import os
+
+# 1. Normalize gene symbols and save normalized 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 the clinical and genetic data
+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 bias and remove biased demographic features
+is_trait_biased, unbiased_linked_data = judge_and_remove_biased_features(linked_data, trait)
+
+# Ensure pure Python bools for JSON serialization
+is_gene_available = bool((normalized_gene_data.shape[0] > 0) and (normalized_gene_data.shape[1] > 0))
+is_trait_available = bool((trait in linked_data.columns) and bool(linked_data[trait].notna().any()))
+is_trait_biased = bool(is_trait_biased)
+
+note = ("INFO: PBMC expression microarray; probes mapped via GENE_SYMBOL; "
+        "no Age feature in clinical annotations; Gender inferred; "
+        "standard GEO pipeline with missingness filtering and imputation.")
+
+# 5. Final validation and save cohort info
+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)

output/preprocess/Type_1_Diabetes/code/GSE162622.py

@@ -0,0 +1,108 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Type_1_Diabetes"
+cohort = "GSE162622"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Type_1_Diabetes"
+in_cohort_dir = "../DATA/GEO/Type_1_Diabetes/GSE162622"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Type_1_Diabetes/GSE162622.csv"
+out_gene_data_file = "./output/z6/preprocess/Type_1_Diabetes/gene_data/GSE162622.csv"
+out_clinical_data_file = "./output/z6/preprocess/Type_1_Diabetes/clinical_data/GSE162622.csv"
+json_path = "./output/z6/preprocess/Type_1_Diabetes/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
+# 1) Gene expression availability
+is_gene_available = True  # Gene expression profiling was performed on UPN119 cells
+
+# 2) Variable availability in sample characteristics
+# Sample Characteristics Dictionary shows only a constant 'cell line' field, no human trait/age/gender.
+trait_row = None
+age_row = None
+gender_row = None
+
+# 2.2) Conversion functions
+def _after_colon(x):
+    if x is None:
+        return None
+    if isinstance(x, str):
+        parts = x.split(":", 1)
+        x = parts[1] if len(parts) > 1 else parts[0]
+    return str(x).strip() if x is not None else None
+
+def convert_trait(x):
+    # No trait information is available for T1D status in this dataset; return None.
+    _ = _after_colon(x)
+    return None
+
+def convert_age(x):
+    v = _after_colon(x)
+    if v is None or v == "" or v.lower() in {"na", "n/a", "unknown"}:
+        return None
+    # Attempt to extract a number (e.g., "35", "35 years")
+    try:
+        return float(''.join(ch for ch in v if (ch.isdigit() or ch == '.' )))
+    except Exception:
+        return None
+
+def convert_gender(x):
+    v = _after_colon(x)
+    if not v:
+        return None
+    vlow = v.lower()
+    if vlow in {"female", "f", "woman", "girl"}:
+        return 0
+    if vlow in {"male", "m", "man", "boy"}:
+        return 1
+    return None
+
+# 3) Save initial 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
+)
+
+# 4) Clinical feature extraction (skip because trait_row is None)
+# If in future a trait_row is identified, uncomment the following block:
+# 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)
+#     os.makedirs(os.path.dirname(out_clinical_data_file), exist_ok=True)
+#     selected_clinical_df.to_csv(out_clinical_data_file)

output/preprocess/Type_1_Diabetes/code/GSE182870.py

@@ -0,0 +1,211 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Type_1_Diabetes"
+cohort = "GSE182870"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Type_1_Diabetes"
+in_cohort_dir = "../DATA/GEO/Type_1_Diabetes/GSE182870"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Type_1_Diabetes/GSE182870.csv"
+out_gene_data_file = "./output/z6/preprocess/Type_1_Diabetes/gene_data/GSE182870.csv"
+out_clinical_data_file = "./output/z6/preprocess/Type_1_Diabetes/clinical_data/GSE182870.csv"
+json_path = "./output/z6/preprocess/Type_1_Diabetes/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
+import re
+import math
+
+# 1) Gene expression data availability (scRNA-seq study)
+is_gene_available = True
+
+# 2) Variable availability based on Sample Characteristics Dictionary
+trait_row = 1   # 'study group: new onset T1D', 'study group: T1D', 'study group: HC'
+age_row = 3     # 'age: <number>'
+gender_row = 4  # 'Sex: male/female'
+
+# 2.2) Conversion functions
+def _after_colon(x):
+    if x is None:
+        return None
+    try:
+        parts = str(x).split(":", 1)
+        return parts[1].strip() if len(parts) > 1 else str(x).strip()
+    except Exception:
+        return None
+
+def convert_trait(x):
+    v = _after_colon(x)
+    if v is None:
+        return None
+    v_low = v.lower()
+    # Map controls
+    if any(k in v_low for k in ["hc", "healthy", "control", "ctrl"]):
+        return 0
+    # Map T1D (including new-onset or established)
+    if any(k in v_low for k in ["t1d", "type 1", "type i", "new onset", "new-onset", "established"]):
+        return 1
+    return None
+
+def convert_age(x):
+    v = _after_colon(x)
+    if v is None:
+        return None
+    m = re.search(r"[-+]?\d*\.?\d+", v)
+    if not m:
+        return None
+    try:
+        val = float(m.group())
+        if abs(val - round(val)) < 1e-6:
+            return int(round(val))
+        return val
+    except Exception:
+        return None
+
+def convert_gender(x):
+    v = _after_colon(x)
+    if v is None:
+        return None
+    v_low = v.lower()
+    if v_low in ["female", "f"]:
+        return 0
+    if v_low in ["male", "m"]:
+        return 1
+    return None
+
+# 3) Save metadata (initial filtering)
+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
+)
+
+# 4) Clinical feature extraction (only if trait data 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 of selected clinical features:", preview)
+    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
+# Step 3: Gene Data Extraction with diagnostics and scRNA-seq fallback
+
+# 1) Try extracting gene expression from the GEO series matrix file
+gene_data = get_genetic_data(matrix_file)
+
+# Diagnostics
+print("Primary extraction (series matrix) gene_data shape:", gene_data.shape)
+
+# If empty, attempt fallback for scRNA-seq supplementary (10x) files
+if gene_data.shape[0] == 0:
+    print("Warning: No expression rows found in the series matrix. Searching for 10x files (matrix.mtx, features/genes.tsv, barcodes.tsv)...")
+    import os
+
+    # Find candidate 10x triplets within the cohort directory
+    tenx_dirs = {}
+    for root, dirs, files in os.walk(in_cohort_dir):
+        files_lower = {f.lower(): f for f in files}
+        has_mtx = 'matrix.mtx' in files_lower or 'matrix.mtx.gz' in files_lower
+        has_features = any(k in files_lower for k in ['features.tsv', 'features.tsv.gz', 'genes.tsv', 'genes.tsv.gz'])
+        has_barcodes = 'barcodes.tsv' in files_lower or 'barcodes.tsv.gz' in files_lower
+        if has_mtx and has_features and has_barcodes:
+            def pick_one(names):
+                for n in names:
+                    if n in files_lower:
+                        return os.path.join(root, files_lower[n])
+                return None
+
+            mtx_path = pick_one(['matrix.mtx', 'matrix.mtx.gz'])
+            features_path = pick_one(['features.tsv', 'features.tsv.gz', 'genes.tsv', 'genes.tsv.gz'])
+            barcodes_path = pick_one(['barcodes.tsv', 'barcodes.tsv.gz'])
+            tenx_dirs[root] = (mtx_path, features_path, barcodes_path)
+
+    if len(tenx_dirs) == 0:
+        print("No 10x triplet found. Gene expression may only be available as external supplementary files not downloaded.")
+    else:
+        # Use the first discovered 10x directory
+        mtx_path, features_path, barcodes_path = list(tenx_dirs.values())[0]
+        print("Found 10x files:")
+        print("  matrix:", mtx_path)
+        print("  features/genes:", features_path)
+        print("  barcodes:", barcodes_path)
+
+        try:
+            import pandas as pd
+            from scipy.io import mmread
+
+            # Read features and barcodes
+            features_df = pd.read_csv(features_path, sep='\t', header=None, compression='infer')
+            barcodes_df = pd.read_csv(barcodes_path, sep='\t', header=None, compression='infer')
+
+            # Determine gene names column (10x v3 features.tsv has: feature_id, gene_name, feature_type)
+            if features_df.shape[1] >= 2:
+                gene_names = features_df.iloc[:, 1].astype(str).tolist()
+            else:
+                gene_names = features_df.iloc[:, 0].astype(str).tolist()
+
+            barcodes = barcodes_df.iloc[:, 0].astype(str).tolist()
+
+            # Read sparse matrix
+            M = mmread(mtx_path)
+
+            # Ensure orientation matches (rows: genes/features; cols: barcodes/cells)
+            if (M.shape[0] == len(gene_names) and M.shape[1] == len(barcodes)):
+                pass
+            elif (M.shape[1] == len(gene_names) and M.shape[0] == len(barcodes)):
+                M = M.T
+            else:
+                print(f"Shape mismatch after reading mtx: {M.shape}, genes: {len(gene_names)}, barcodes: {len(barcodes)}")
+                M = None
+
+            if M is not None:
+                # Build a sparse DataFrame to avoid densification
+                gene_data = pd.DataFrame.sparse.from_spmatrix(M, index=gene_names, columns=barcodes)
+                print("10x extraction gene_data shape:", gene_data.shape)
+        except ImportError as e:
+            print("SciPy is required to read matrix.mtx but is not available. Skipping 10x fallback.")
+        except Exception as e:
+            print("An error occurred while reading 10x files:", e)
+
+# 2) Print the first 20 row IDs (gene or probe identifiers)
+try:
+    print("First 20 row IDs:", list(gene_data.index[:20]))
+except Exception as e:
+    print("Unable to print row IDs. gene_data may be undefined or malformed:", e)

output/preprocess/Type_1_Diabetes/code/GSE193273.py

@@ -0,0 +1,193 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Type_1_Diabetes"
+cohort = "GSE193273"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Type_1_Diabetes"
+in_cohort_dir = "../DATA/GEO/Type_1_Diabetes/GSE193273"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Type_1_Diabetes/GSE193273.csv"
+out_gene_data_file = "./output/z6/preprocess/Type_1_Diabetes/gene_data/GSE193273.csv"
+out_clinical_data_file = "./output/z6/preprocess/Type_1_Diabetes/clinical_data/GSE193273.csv"
+json_path = "./output/z6/preprocess/Type_1_Diabetes/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
+# Determine data availability
+is_gene_available = True  # Whole genome expression on PBMC (not miRNA-only or methylation)
+trait_row = 1  # 'diagnosis' field
+age_row = None
+gender_row = None
+
+# Conversion functions
+def _after_colon(x):
+    if x is None:
+        return None
+    s = str(x)
+    parts = s.split(":", 1)
+    return parts[1].strip() if len(parts) > 1 else s.strip()
+
+def convert_trait(x):
+    val = _after_colon(x)
+    if val is None:
+        return None
+    v = val.lower()
+    # Map controls to 0
+    if "control" in v:
+        return 0
+    # Map T1D cases to 1
+    if ("t1d" in v) or ("type 1" in v) or ("type i" in v) or ("type-1" in v) or ("diabetes" in v):
+        return 1
+    return None
+
+def convert_age(x):
+    val = _after_colon(x)
+    if val is None:
+        return None
+    # Extract the first numeric token as age (years)
+    import re
+    m = re.search(r"[-+]?\d*\.?\d+", val)
+    if m:
+        try:
+            return float(m.group())
+        except Exception:
+            return None
+    return None
+
+def convert_gender(x):
+    val = _after_colon(x)
+    if val is None:
+        return None
+    v = val.lower()
+    if v in {"female", "f", "woman", "women"}:
+        return 0
+    if v in {"male", "m", "man", "men"}:
+        return 1
+    return None
+
+# Initial metadata recording
+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
+)
+
+# Clinical feature extraction and saving
+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)
+
+    import os
+    os.makedirs(os.path.dirname(out_clinical_data_file), exist_ok=True)
+    selected_clinical_df.to_csv(out_clinical_data_file)
+
+# 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
+# The observed identifiers are numeric strings like '4', '5', ... '23', which are not human gene symbols.
+requires_gene_mapping = True
+print(f"requires_gene_mapping = {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
+# Identify the appropriate columns in the annotation
+probe_id_col = 'ID'           # Matches the numeric feature IDs seen in the matrix (e.g., '4', '5', ...)
+gene_symbol_col = 'GENE_SYMBOL'
+
+# Ensure probe-level expression data is available
+try:
+    gene_data
+except NameError:
+    gene_data = get_genetic_data(matrix_file)
+
+# Build mapping and apply it to convert probe-level data to gene-level data
+mapping_df = get_gene_mapping(gene_annotation, prob_col=probe_id_col, gene_col=gene_symbol_col)
+gene_data = apply_gene_mapping(expression_df=gene_data, mapping_df=mapping_df)
+
+# Step 7: Data Normalization and Linking
+import os
+
+# 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 check and remove biased demographic features
+is_trait_biased, unbiased_linked_data = judge_and_remove_biased_features(linked_data, trait)
+is_trait_biased = bool(is_trait_biased)
+
+# Dynamically infer availability flags for final validation (cast to built-in bool)
+is_gene_available = bool((normalized_gene_data.shape[0] > 0) and (normalized_gene_data.shape[1] > 0))
+is_trait_available = bool((trait in selected_clinical_df.index) and selected_clinical_df.loc[trait].notna().any())
+
+# 5. Final validation and save cohort info
+note = "INFO: Only diagnosis (trait) available; Age/Gender not provided in series characteristics."
+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)

output/preprocess/Type_1_Diabetes/code/GSE232310.py

@@ -0,0 +1,155 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Type_1_Diabetes"
+cohort = "GSE232310"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Type_1_Diabetes"
+in_cohort_dir = "../DATA/GEO/Type_1_Diabetes/GSE232310"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Type_1_Diabetes/GSE232310.csv"
+out_gene_data_file = "./output/z6/preprocess/Type_1_Diabetes/gene_data/GSE232310.csv"
+out_clinical_data_file = "./output/z6/preprocess/Type_1_Diabetes/clinical_data/GSE232310.csv"
+json_path = "./output/z6/preprocess/Type_1_Diabetes/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
+
+# 1) Gene expression data availability
+is_gene_available = True  # RNA extracted from monocytes -> gene expression data likely available
+
+# 2) Variable availability and conversion functions
+trait_row = 1
+age_row = None
+gender_row = None
+
+def convert_trait(x):
+    if x is None:
+        return None
+    # Extract value after colon
+    try:
+        val = str(x).split(":", 1)[1].strip().lower()
+    except Exception:
+        val = str(x).strip().lower()
+    # Map to binary: ROT1D = 1 (case), others (HRS, LRS, uHC) = 0 (controls)
+    if "rot1d" in val or "recent on" in val:  # capture 'recent onset' variants
+        return 1
+    if any(k in val for k in ["uhc", "hrs", "lrs", "healthy", "sibling", "control"]):
+        return 0
+    return None
+
+convert_age = None
+convert_gender = None
+
+# 3) Save metadata (initial filtering)
+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
+)
+
+# 4) Clinical feature extraction (only if clinical 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
+    )
+    print(preview_df(selected_clinical_df))
+    os.makedirs(os.path.dirname(out_clinical_data_file), exist_ok=True)
+    selected_clinical_df.to_csv(out_clinical_data_file)
+
+# 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
+requires_gene_mapping = True
+print(f"requires_gene_mapping = {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
+# Determine appropriate columns for probe IDs and gene symbols based on preview:
+# Probe identifier column: 'ID' (e.g., '1007_s_at')
+# Gene symbol column: 'Gene Symbol' (e.g., 'DDR1 /// MIR4640')
+
+# 1-2) Build mapping dataframe from annotation
+mapping_df = get_gene_mapping(gene_annotation, prob_col='ID', gene_col='Gene Symbol')
+
+# 3) Apply mapping to convert probe-level data to gene-level expression
+gene_data = apply_gene_mapping(expression_df=gene_data, mapping_df=mapping_df)
+
+# Step 7: Data Normalization and Linking
+import os
+
+# 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. Assess bias and remove biased demographic features
+is_trait_biased, unbiased_linked_data = judge_and_remove_biased_features(linked_data, trait)
+
+# 5. Final validation and save cohort info
+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,
+    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)

output/preprocess/Type_1_Diabetes/code/GSE71799.py

@@ -0,0 +1,133 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Type_1_Diabetes"
+cohort = "GSE71799"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Type_1_Diabetes"
+in_cohort_dir = "../DATA/GEO/Type_1_Diabetes/GSE71799"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Type_1_Diabetes/GSE71799.csv"
+out_gene_data_file = "./output/z6/preprocess/Type_1_Diabetes/gene_data/GSE71799.csv"
+out_clinical_data_file = "./output/z6/preprocess/Type_1_Diabetes/clinical_data/GSE71799.csv"
+json_path = "./output/z6/preprocess/Type_1_Diabetes/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
+# Step 1: Assess data availability based on provided background and sample characteristics
+is_gene_available = True  # Gene expression profiling was performed (not miRNA/methylation)
+trait_row = None          # No Type 1 Diabetes status recorded; cohort compares CF vs healthy controls
+age_row = None            # No age information available
+gender_row = None         # No gender information available
+
+# Step 2: Define conversion functions
+def _after_colon(x):
+    if x is None:
+        return None
+    s = str(x)
+    parts = s.split(":", 1)
+    val = parts[1] if len(parts) == 2 else parts[0]
+    return val.strip().strip('"').strip("'")
+
+def convert_trait(x):
+    # Binary: 1 = Type 1 Diabetes, 0 = non-T1D
+    v = _after_colon(x)
+    if v is None or v == "":
+        return None
+    v_lower = v.lower()
+    t1d_pos = [
+        "type 1 diabetes", "t1d", "t1dm", "recent-onset t1d", "recent onset t1d",
+        "recent-onset type 1 diabetes", "recent onset type 1 diabetes", "ro t1d",
+        "type i diabetes", "diabetic", "type 1 dm"
+    ]
+    t1d_neg = [
+        "control", "healthy", "hc", "uhc", "unrelated healthy control", "non-diabetic",
+        "no diabetes", "cystic fibrosis", "cf"
+    ]
+    if any(k in v_lower for k in t1d_pos):
+        return 1
+    if any(k in v_lower for k in t1d_neg):
+        return 0
+    return None
+
+def convert_age(x):
+    # Continuous: years
+    v = _after_colon(x)
+    if v is None or v == "":
+        return None
+    s = v.lower()
+    # Extract first float-like number
+    import re
+    m = re.search(r"[-+]?\d*\.?\d+", s)
+    if not m:
+        return None
+    num = float(m.group())
+    if "month" in s:
+        return num / 12.0
+    if "week" in s:
+        return num / 52.0
+    if "day" in s:
+        return num / 365.0
+    # default assume years
+    return num
+
+def convert_gender(x):
+    # Binary: female -> 0, male -> 1
+    v = _after_colon(x)
+    if v is None or v == "":
+        return None
+    s = v.strip().lower()
+    if s in {"female", "f", "woman", "girl"}:
+        return 0
+    if s in {"male", "m", "man", "boy"}:
+        return 1
+    return None
+
+# Step 3: Initial filtering 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 (skipped because trait_row is None)
+# If in another scenario trait_row is not None, the following block would run:
+if trait_row is not None:
+    selected_clinical = 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, n=5)
+    os.makedirs(os.path.dirname(out_clinical_data_file), exist_ok=True)
+    selected_clinical.to_csv(out_clinical_data_file, index=True)

output/preprocess/Type_1_Diabetes/code/GSE75062.py

@@ -0,0 +1,158 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Type_1_Diabetes"
+cohort = "GSE75062"
+
+# Input paths
+in_trait_dir = "../DATA/GEO/Type_1_Diabetes"
+in_cohort_dir = "../DATA/GEO/Type_1_Diabetes/GSE75062"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Type_1_Diabetes/GSE75062.csv"
+out_gene_data_file = "./output/z6/preprocess/Type_1_Diabetes/gene_data/GSE75062.csv"
+out_clinical_data_file = "./output/z6/preprocess/Type_1_Diabetes/clinical_data/GSE75062.csv"
+json_path = "./output/z6/preprocess/Type_1_Diabetes/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 re
+
+# 1) Gene expression data availability (based on series description: microarray gene expression of human islets)
+is_gene_available = True
+
+# 2) Variable availability (from Sample Characteristics Dictionary)
+# Keys present: 
+# 0 -> 'diabetes reversal status: Yes/No' (mouse outcome, not human T1D status)
+# 1 -> 'tissue: pancreas' (constant)
+# 2 -> 'cell type: islet cells' (constant)
+trait_row = None      # No human Type 1 Diabetes status available
+age_row = None        # No age information available
+gender_row = None     # No gender information available
+
+# 2.2) Converters
+def _after_colon(x: str) -> str:
+    if x is None:
+        return ""
+    parts = str(x).split(':', 1)
+    return parts[1].strip() if len(parts) > 1 else str(x).strip()
+
+def convert_trait(x):
+    # Binary: 1 = Type 1 Diabetes, 0 = Not Type 1 Diabetes
+    v = _after_colon(x).lower()
+    if not v:
+        return None
+    # Positive indicators for T1D
+    if any(k in v for k in ["type 1 diabetes", "type i diabetes", "t1d", "t1dm"]):
+        return 1
+    # Negative indicators (explicitly not T1D)
+    if any(k in v for k in ["non-diabetic", "nondiabetic", "control", "healthy", "normoglycemic", "no diabetes", "type 2 diabetes", "t2d", "t2dm"]):
+        return 0
+    # If the field is about "diabetes reversal status", it's not human T1D; avoid misclassification
+    if "reversal" in x.lower():
+        return None
+    return None
+
+def convert_age(x):
+    v = _after_colon(x)
+    if not v:
+        return None
+    # Extract first numeric value (years)
+    m = re.search(r"[-+]?\d*\.?\d+", v)
+    if not m:
+        return None
+    try:
+        age = float(m.group())
+        if 0 < age < 120:
+            return age
+    except Exception:
+        pass
+    return None
+
+def convert_gender(x):
+    v = _after_colon(x).lower()
+    if not v:
+        return None
+    if v in ["female", "f", "woman", "women", "girl"]:
+        return 0
+    if v in ["male", "m", "man", "men", "boy"]:
+        return 1
+    return None
+
+# 3) Save metadata (initial filtering)
+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
+)
+
+# 4) Clinical feature extraction (skip because trait_row is None)
+if trait_row is not None:
+    selected = 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, n=5)
+    selected.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
+requires_gene_mapping = True
+print(f"requires_gene_mapping = {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
+# Identify the appropriate columns for probe IDs and gene symbols based on the annotation preview:
+# Probe identifiers: 'ID'
+# Gene symbols: 'Gene Symbol'
+
+# Preserve the original probe-level data
+probe_data = gene_data
+
+# 2. Get mapping dataframe
+mapping_df = get_gene_mapping(gene_annotation, prob_col='ID', gene_col='Gene Symbol')
+
+# 3. Apply mapping to convert probe-level data to gene-level expression
+gene_data = apply_gene_mapping(probe_data, mapping_df)

output/preprocess/Type_1_Diabetes/code/TCGA.py

@@ -0,0 +1,78 @@
+# Path Configuration
+from tools.preprocess import *
+
+# Processing context
+trait = "Type_1_Diabetes"
+
+# Input paths
+tcga_root_dir = "../DATA/TCGA"
+
+# Output paths
+out_data_file = "./output/z6/preprocess/Type_1_Diabetes/TCGA.csv"
+out_gene_data_file = "./output/z6/preprocess/Type_1_Diabetes/gene_data/TCGA.csv"
+out_clinical_data_file = "./output/z6/preprocess/Type_1_Diabetes/clinical_data/TCGA.csv"
+json_path = "./output/z6/preprocess/Type_1_Diabetes/cohort_info.json"
+
+
+# Step 1: Initial Data Loading
+import os
+import pandas as pd
+
+# Given TCGA subdirectories (from the task)
+tcga_subdirs = [
+    'TCGA_lower_grade_glioma_and_glioblastoma_(GBMLGG)', 'TCGA_Uterine_Carcinosarcoma_(UCS)',
+    'TCGA_Thyroid_Cancer_(THCA)', 'TCGA_Thymoma_(THYM)', 'TCGA_Testicular_Cancer_(TGCT)',
+    'TCGA_Stomach_Cancer_(STAD)', 'TCGA_Sarcoma_(SARC)', 'TCGA_Rectal_Cancer_(READ)',
+    'TCGA_Prostate_Cancer_(PRAD)', 'TCGA_Pheochromocytoma_Paraganglioma_(PCPG)',
+    'TCGA_Pancreatic_Cancer_(PAAD)', 'TCGA_Ovarian_Cancer_(OV)', 'TCGA_Ocular_melanomas_(UVM)',
+    'TCGA_Mesothelioma_(MESO)', 'TCGA_Melanoma_(SKCM)', 'TCGA_Lung_Squamous_Cell_Carcinoma_(LUSC)',
+    'TCGA_Lung_Cancer_(LUNG)', 'TCGA_Lung_Adenocarcinoma_(LUAD)', 'TCGA_Lower_Grade_Glioma_(LGG)',
+    'TCGA_Liver_Cancer_(LIHC)', 'TCGA_Large_Bcell_Lymphoma_(DLBC)', 'TCGA_Kidney_Papillary_Cell_Carcinoma_(KIRP)',
+    'TCGA_Kidney_Clear_Cell_Carcinoma_(KIRC)', 'TCGA_Kidney_Chromophobe_(KICH)',
+    'TCGA_Head_and_Neck_Cancer_(HNSC)', 'TCGA_Glioblastoma_(GBM)', 'TCGA_Esophageal_Cancer_(ESCA)',
+    'TCGA_Endometrioid_Cancer_(UCEC)', 'TCGA_Colon_and_Rectal_Cancer_(COADREAD)',
+    'TCGA_Colon_Cancer_(COAD)', 'TCGA_Cervical_Cancer_(CESC)', 'TCGA_Breast_Cancer_(BRCA)',
+    'TCGA_Bladder_Cancer_(BLCA)', 'TCGA_Bile_Duct_Cancer_(CHOL)', 'TCGA_Adrenocortical_Cancer_(ACC)',
+    'TCGA_Acute_Myeloid_Leukemia_(LAML)'
+]
+
+# Try to find a cohort related to Type 1 Diabetes (unlikely in TCGA, which is cancer-focused)
+synonyms = {
+    "type_1_diabetes", "type 1 diabetes", "t1d", "insulin-dependent", "insulin dependent", "juvenile diabetes",
+    "diabetes"
+}
+selected_dir = None
+for d in tcga_subdirs:
+    name = d.lower()
+    if any(term in name for term in synonyms):
+        selected_dir = d
+        break
+
+clinical_df = None
+genetic_df = None
+clinical_file_path = None
+genetic_file_path = None
+skip_tcga_processing = False
+
+if selected_dir is None:
+    print("No suitable TCGA cohort matching the trait 'Type_1_Diabetes' was found. Skipping TCGA for this trait.")
+    # Record metadata: no suitable cohort, thus no gene/trait available for this trait in TCGA
+    _ = validate_and_save_cohort_info(
+        is_final=False,
+        cohort="TCGA",
+        info_path=json_path,
+        is_gene_available=False,
+        is_trait_available=False
+    )
+    skip_tcga_processing = True
+else:
+    # Identify key file paths within the selected cohort directory
+    cohort_dir = os.path.join(tcga_root_dir, selected_dir)
+    clinical_file_path, genetic_file_path = tcga_get_relevant_filepaths(cohort_dir)
+
+    # Load the clinical and genetic data
+    clinical_df = pd.read_csv(clinical_file_path, sep='\t', index_col=0, low_memory=False)
+    genetic_df = pd.read_csv(genetic_file_path, sep='\t', index_col=0, low_memory=False)
+
+    # Print clinical column names
+    print(list(clinical_df.columns))

output/preprocess/Type_1_Diabetes/cohort_info.json

@@ -1,112 +1 @@
-{
-  "GSE156035": {
-    "is_usable": true,
-    "is_gene_available": true,
-    "is_trait_available": true,
-    "is_available": true,
-    "is_biased": false,
-    "has_age": false,
-    "has_gender": true,
-    "sample_size": 40
-  },
-  "GSE75062": {
-    "is_usable": true,
-    "is_gene_available": true,
-    "is_trait_available": true,
-    "is_available": true,
-    "is_biased": false,
-    "has_age": false,
-    "has_gender": false,
-    "sample_size": 59
-  },
-  "GSE71799": {
-    "is_usable": false,
-    "is_gene_available": false,
-    "is_trait_available": false,
-    "is_available": false,
-    "is_biased": null,
-    "has_age": null,
-    "has_gender": null,
-    "sample_size": null
-  },
-  "GSE232310": {
-    "is_usable": true,
-    "is_gene_available": true,
-    "is_trait_available": true,
-    "is_available": true,
-    "is_biased": false,
-    "has_age": false,
-    "has_gender": false,
-    "sample_size": 62
-  },
-  "GSE193273": {
-    "is_usable": true,
-    "is_gene_available": true,
-    "is_trait_available": true,
-    "is_available": true,
-    "is_biased": false,
-    "has_age": false,
-    "has_gender": false,
-    "sample_size": 2
-  },
-  "GSE182870": {
-    "is_usable": false,
-    "is_gene_available": false,
-    "is_trait_available": true,
-    "is_available": false,
-    "is_biased": null,
-    "has_age": null,
-    "has_gender": null,
-    "sample_size": null
-  },
-  "GSE162622": {
-    "is_usable": false,
-    "is_gene_available": true,
-    "is_trait_available": false,
-    "is_available": false,
-    "is_biased": null,
-    "has_age": null,
-    "has_gender": null,
-    "sample_size": null
-  },
-  "GSE131528": {
-    "is_usable": false,
-    "is_gene_available": true,
-    "is_trait_available": true,
-    "is_available": true,
-    "is_biased": true,
-    "has_age": true,
-    "has_gender": false,
-    "sample_size": 55
-  },
-  "GSE123088": {
-    "is_usable": false,
-    "is_gene_available": false,
-    "is_trait_available": false,
-    "is_available": false,
-    "is_biased": null,
-    "has_age": null,
-    "has_gender": null,
-    "sample_size": null
-  },
-  "GSE123086": {
-    "is_usable": false,
-    "is_gene_available": false,
-    "is_trait_available": false,
-    "is_available": false,
-    "is_biased": null,
-    "has_age": null,
-    "has_gender": null,
-    "sample_size": null
-  },
-  "TCGA": {
-    "is_usable": false,
-    "is_gene_available": false,
-    "is_trait_available": false,
-    "is_available": false,
-    "is_biased": null,
-    "has_age": null,
-    "has_gender": null,
-    "sample_size": null
-  }
-}
+{"GSE75062": {"is_usable": false, "is_gene_available": true, "is_trait_available": false, "is_available": false, "is_biased": null, "has_age": null, "has_gender": null, "sample_size": null, "note": null}, "GSE71799": {"is_usable": false, "is_gene_available": true, "is_trait_available": false, "is_available": false, "is_biased": null, "has_age": null, "has_gender": null, "sample_size": null, "note": null}, "GSE232310": {"is_usable": true, "is_gene_available": true, "is_trait_available": true, "is_available": true, "is_biased": false, "has_age": false, "has_gender": false, "sample_size": 62, "note": ""}, "GSE193273": {"is_usable": true, "is_gene_available": true, "is_trait_available": true, "is_available": true, "is_biased": false, "has_age": false, "has_gender": false, "sample_size": 40, "note": "INFO: Only diagnosis (trait) available; Age/Gender not provided in series characteristics."}, "GSE162622": {"is_usable": false, "is_gene_available": true, "is_trait_available": false, "is_available": false, "is_biased": null, "has_age": null, "has_gender": null, "sample_size": null, "note": null}, "GSE156035": {"is_usable": true, "is_gene_available": true, "is_trait_available": true, "is_available": true, "is_biased": false, "has_age": false, "has_gender": true, "sample_size": 40, "note": "INFO: PBMC expression microarray; probes mapped via GENE_SYMBOL; no Age feature in clinical annotations; Gender inferred; standard GEO pipeline with missingness filtering and imputation."}, "GSE131528": {"is_usable": false, "is_gene_available": true, "is_trait_available": false, "is_available": false, "is_biased": null, "has_age": null, "has_gender": null, "sample_size": null, "note": null}, "GSE123088": {"is_usable": true, "is_gene_available": true, "is_trait_available": true, "is_available": true, "is_biased": false, "has_age": true, "has_gender": true, "sample_size": 204, "note": "INFO: Gene normalization skipped because gene IDs are Entrez (ENTREZ:####)."}, "GSE123086": {"is_usable": true, "is_gene_available": true, "is_trait_available": true, "is_available": true, "is_biased": false, "has_age": true, "has_gender": true, "sample_size": 47, "note": "WARNING: Skipped gene-symbol normalization because gene index consists of Entrez IDs and no gene symbol annotations were available."}, "TCGA": {"is_usable": false, "is_gene_available": false, "is_trait_available": false, "is_available": false, "is_biased": null, "has_age": null, "has_gender": null, "sample_size": null, "note": null}}