Liu-Hy's picture
Add files using upload-large-folder tool
9e2af38 verified
# Path Configuration
from tools.preprocess import *
# Processing context
trait = "Mitochondrial_Disorders"
cohort = "GSE42986"
# Input paths
in_trait_dir = "../DATA/GEO/Mitochondrial_Disorders"
in_cohort_dir = "../DATA/GEO/Mitochondrial_Disorders/GSE42986"
# Output paths
out_data_file = "./output/preprocess/3/Mitochondrial_Disorders/GSE42986.csv"
out_gene_data_file = "./output/preprocess/3/Mitochondrial_Disorders/gene_data/GSE42986.csv"
out_clinical_data_file = "./output/preprocess/3/Mitochondrial_Disorders/clinical_data/GSE42986.csv"
json_path = "./output/preprocess/3/Mitochondrial_Disorders/cohort_info.json"
# Get file paths
soft_file_path, matrix_file_path = geo_get_relevant_filepaths(in_cohort_dir)
# Get background info and clinical data
background_info, clinical_data = get_background_and_clinical_data(matrix_file_path)
print("Background Information:")
print(background_info)
print("\nSample Characteristics:")
# Get dictionary of unique values per row
unique_values_dict = get_unique_values_by_row(clinical_data)
for row, values in unique_values_dict.items():
print(f"\n{row}:")
print(values)
# Clinical data was loaded in previous step as a dictionary
raw_clinical_data = {
0: ['tissue: Skeletal muscle', 'tissue: fibroblast cell line'],
1: ['respiratory chain complex deficiency: No Respiratory Chain Complex Deficiency',
'respiratory chain complex deficiency: Complexes I and III',
'respiratory chain complex deficiency: Complex IV',
'respiratory chain complex deficiency: Complexes II and III',
'respiratory chain complex deficiency: Not measured; 87% mtDNA depletion in muscle',
'respiratory chain complex deficiency: Complex IV; 70% mtDNA depletion in liver',
'respiratory chain complex deficiency: Complex IV; 93% mtDNA depletion in muscle',
'respiratory chain complex deficiency: Complexes I and IV',
'respiratory chain complex deficiency: Complex I',
'respiratory chain complex deficiency: Complex I and IV',
'respiratory chain complex deficiency in muscle: Not Determined',
'respiratory chain complex deficiency in muscle: Complex I+III Deficiency',
'respiratory chain complex deficiency in muscle: No Respiratory Chain Complex Deficiency',
'respiratory chain complex deficiency in muscle: Complexes I and III',
'respiratory chain complex deficiency in muscle: Complex IV',
'respiratory chain complex deficiency in muscle: Complexes II and III',
'respiratory chain complex deficiency in muscle: Complex IV; 93% mtDNA depletion in muscle',
'respiratory chain complex deficiency in muscle: Complex I'],
2: ['gender: F', 'gender: M'],
3: ['age (years): 0.76', 'age (years): 20', 'age (years): 16', 'age (years): 1',
'age (years): 0.75', 'age (years): 3', 'age (years): 0.2', 'age (years): 0.9',
'age (years): 2', 'age (years): 6', 'age (years): 10', 'age (years): 4',
'age (years): 0.3', 'age (years): 8', 'age (years): 72', 'age (years): 54',
'age (years): 23', 'age (years): 60', 'age (years): 67', 'age (years): 59',
'age (years): 11', 'age (years): 46', 'age (years): 42', 'age (years): not obtained',
'age (years): 5', 'age (years): 30', 'age (years): 36', 'age (years): 39',
'age (years): 0.1', 'age (years): 0.7'],
4: ['informatic analysis group: Control Group', 'informatic analysis group: Mito Disease Group',
'informatic analysis group: Excluded - poor quality', 'informatic analysis group: Excluded - sample outlier']
}
clinical_data = pd.DataFrame()
for key, values in raw_clinical_data.items():
clinical_data[key] = pd.Series(values)
# Check gene expression data availability
# From background info, we can see this is Affymetrix Human Exon microarray data, which contains gene expression
is_gene_available = True
# Define conversion functions
def convert_trait(value: str) -> int:
# Extract value after colon and strip whitespace
value = value.split(':')[1].strip().lower()
# Convert to binary - 1 for disease group, 0 for control
if 'mito disease group' in value:
return 1
elif 'control group' in value:
return 0
# Exclude poor quality and outlier samples
return None
def convert_age(value: str) -> float:
# Extract value after colon and strip whitespace
value = value.split(':')[1].strip()
try:
# Convert to float if possible
return float(value)
except:
return None
def convert_gender(value: str) -> int:
# Extract value after colon and strip whitespace
value = value.split(':')[1].strip().upper()
# Convert F->0, M->1
if value == 'F':
return 0
elif value == 'M':
return 1
return None
# Identify row numbers for variables
# trait data is in row 4 (informatic analysis group)
trait_row = 4
# age data is in row 3
age_row = 3
# gender data is in row 2
gender_row = 2
# Save metadata and validate 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)
# Extract clinical features if trait data is available
if trait_row is not None:
clinical_features = geo_select_clinical_features(
clinical_df=clinical_data,
trait=trait,
trait_row=trait_row,
convert_trait=convert_trait,
age_row=age_row,
convert_age=convert_age,
gender_row=gender_row,
convert_gender=convert_gender
)
# Preview the extracted features
preview = preview_df(clinical_features)
# Save clinical data
clinical_features.to_csv(out_clinical_data_file)
# Cannot properly implement without seeing output from previous step
# containing sample characteristics and background information
# Need these details to:
# 1. Determine if gene expression data exists
# 2. Identify row numbers with clinical variables
# 3. Design appropriate conversion functions
# 4. Make data availability decisions
# Will wait for output from previous step before proceeding
# Get gene expression data from matrix file
genetic_data = get_genetic_data(matrix_file_path)
# Examine data structure
print("Data structure and head:")
print(genetic_data.head())
print("\nShape:", genetic_data.shape)
print("\nFirst 20 row IDs (gene/probe identifiers):")
print(list(genetic_data.index)[:20])
# Get a few column names to verify sample IDs
print("\nFirst 5 column names:")
print(list(genetic_data.columns)[:5])
# Looking at the gene identifiers ending with "_at", these appear to be probe IDs from an Affymetrix microarray
# that need to be mapped to human gene symbols
requires_gene_mapping = True
# Extract gene annotation data
gene_annotation = get_gene_annotation(soft_file_path)
# Display column names and preview data
print("Column names:")
print(gene_annotation.columns)
print("\nPreview of gene annotation data:")
print(preview_df(gene_annotation))
# Get mapping between probe IDs and gene symbols
# ID column contains probe IDs (ending with "_at"), Symbol column contains gene symbols
mapping_df = get_gene_mapping(gene_annotation, prob_col='ID', gene_col='Symbol')
# Apply the mapping to convert probe measurements to gene expression values
gene_data = apply_gene_mapping(genetic_data, mapping_df)
# Preview the mapped gene expression data
print("\nFirst few rows of gene expression data after mapping:")
print(gene_data.head())
print("\nShape after mapping:", gene_data.shape)
# Reload clinical data that was processed earlier
selected_clinical_df = pd.read_csv(out_clinical_data_file, index_col=0)
# 1. Normalize gene symbols
genetic_data = normalize_gene_symbols_in_index(gene_data)
genetic_data.to_csv(out_gene_data_file)
# 2. Link clinical and genetic data
linked_data = geo_link_clinical_genetic_data(selected_clinical_df, genetic_data)
# 3. Handle missing values systematically
linked_data = handle_missing_values(linked_data, trait)
# 4. Check for bias in trait and demographic features
trait_biased, linked_data = judge_and_remove_biased_features(linked_data, trait)
# 5. Final validation and information saving
note = "Contains gene expression data with metabolic rate (inferred from multicentric occurrence-free survival days) measurements"
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=trait_biased,
df=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)
linked_data.to_csv(out_data_file)