Liu-Hy's picture
Add files using upload-large-folder tool
d5514d2 verified
# Path Configuration
from tools.preprocess import *
# Processing context
trait = "Rheumatoid_Arthritis"
cohort = "GSE224330"
# Input paths
in_trait_dir = "../DATA/GEO/Rheumatoid_Arthritis"
in_cohort_dir = "../DATA/GEO/Rheumatoid_Arthritis/GSE224330"
# Output paths
out_data_file = "./output/preprocess/3/Rheumatoid_Arthritis/GSE224330.csv"
out_gene_data_file = "./output/preprocess/3/Rheumatoid_Arthritis/gene_data/GSE224330.csv"
out_clinical_data_file = "./output/preprocess/3/Rheumatoid_Arthritis/clinical_data/GSE224330.csv"
json_path = "./output/preprocess/3/Rheumatoid_Arthritis/cohort_info.json"
# Get file paths
soft_file, matrix_file = geo_get_relevant_filepaths(in_cohort_dir)
# Extract background info and clinical data
background_info, clinical_data = get_background_and_clinical_data(matrix_file)
# Get unique values per clinical feature
sample_characteristics = get_unique_values_by_row(clinical_data)
# Print background info
print("Dataset Background Information:")
print(f"{background_info}\n")
# Print sample characteristics
print("Sample Characteristics:")
for feature, values in sample_characteristics.items():
print(f"Feature: {feature}")
print(f"Values: {values}\n")
# 1. Gene Expression Data Availability
# Based on background info mentioning "gene expression profiling", "transcriptomic profile", "whole-genome transcriptomics"
is_gene_available = True
# 2.1 Variable Availability
trait_row = 0 # Can infer RA status from tissue source
age_row = 1 # Age data available in feature 1
gender_row = 2 # Gender data available in feature 2
# 2.2 Data Type Conversion Functions
def convert_trait(x):
if pd.isna(x):
return None
# First 10 samples (GSM7019507-GSM7019516) are from healthy controls based on background info
# Rest are RA patients on different treatments
sample_id = x.name
sample_num = int(sample_id.replace('GSM',''))
if 7019507 <= sample_num <= 7019516:
return 0 # Healthy control
else:
return 1 # RA patient
def convert_age(x):
if pd.isna(x):
return None
# Extract numeric value before 'y'
try:
age = int(x.split(':')[1].strip().replace('y',''))
return age
except:
return None
def convert_gender(x):
if pd.isna(x):
return None
value = x.split(':')[1].strip().lower()
if 'female' in value:
return 0
elif 'male' in value:
return 1
return None
# 3. 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
selected_clinical_df = geo_select_clinical_features(
clinical_df=clinical_data,
trait=trait,
trait_row=trait_row,
convert_trait=convert_trait,
age_row=age_row,
convert_age=convert_age,
gender_row=gender_row,
convert_gender=convert_gender
)
# Preview the extracted features
preview = preview_df(selected_clinical_df)
print("Preview of extracted clinical features:")
print(preview)
# Save to CSV
selected_clinical_df.to_csv(out_clinical_data_file)
# The previous step output was not provided. Without it, we cannot properly:
# 1. Determine gene expression data availability
# 2. Identify row numbers for clinical features
# 3. Design appropriate conversion logic based on actual data values
# Therefore, this step cannot be completed until we receive:
# - Background information about the dataset
# - Sample characteristics dictionary showing available clinical data
raise ValueError("Previous step output with dataset information is required to analyze data availability and implement conversion logic")
# Get file paths
soft_file, matrix_file = geo_get_relevant_filepaths(in_cohort_dir)
# Extract gene expression data from matrix file
gene_data = get_genetic_data(matrix_file)
# Print first 20 row IDs and shape of data to help debug
print("Shape of gene expression data:", gene_data.shape)
print("\nFirst few rows of data:")
print(gene_data.head())
print("\nFirst 20 gene/probe identifiers:")
print(gene_data.index[:20])
# Inspect a snippet of raw file to verify identifier format
import gzip
with gzip.open(matrix_file, 'rt', encoding='utf-8') as f:
lines = []
for i, line in enumerate(f):
if "!series_matrix_table_begin" in line:
# Get the next 5 lines after the marker
for _ in range(5):
lines.append(next(f).strip())
break
print("\nFirst few lines after matrix marker in raw file:")
for line in lines:
print(line)
# The identifiers starting with "A_19_P" appear to be Agilent microarray probe IDs
# These are not standard human gene symbols and need to be mapped to gene symbols
requires_gene_mapping = True
# Extract gene annotation data
gene_metadata = get_gene_annotation(soft_file)
# Preview the annotation data
print("Column names:", gene_metadata.columns.tolist())
print("\nFirst few rows preview:")
print(preview_df(gene_metadata))
# 1. Extract gene annotation data
gene_metadata = get_gene_annotation(soft_file)
# 2. Extract gene mapping from annotation data
mapping_data = get_gene_mapping(gene_metadata, prob_col='ID', gene_col='GENE_SYMBOL')
# 3. Apply mapping to convert probe-level data to gene-level data
gene_expression_data = apply_gene_mapping(expression_df=gene_data, mapping_df=mapping_data)
# Save processed gene data
gene_expression_data.to_csv(out_gene_data_file)
# Get file paths
soft_file, matrix_file = geo_get_relevant_filepaths(in_cohort_dir)
# Extract gene expression data from matrix file
gene_data = get_genetic_data(matrix_file)
# Print first 20 row IDs and shape of data to help debug
print("Shape of gene expression data:", gene_data.shape)
print("\nFirst few rows of data:")
print(gene_data.head())
print("\nFirst 20 gene/probe identifiers:")
print(gene_data.index[:20])
# Inspect a snippet of raw file to verify identifier format
import gzip
with gzip.open(matrix_file, 'rt', encoding='utf-8') as f:
lines = []
for i, line in enumerate(f):
if "!series_matrix_table_begin" in line:
# Get the next 5 lines after the marker
for _ in range(5):
lines.append(next(f).strip())
break
print("\nFirst few lines after matrix marker in raw file:")
for line in lines:
print(line)
# 1. Extract gene annotation data and observe identifiers
# From previous outputs, we can see:
# - Gene expression data uses identifiers like 'A_19_P00315452'
# - Gene annotation data has matching IDs in the 'ID' column and gene symbols in 'GENE_SYMBOL'
gene_metadata = get_gene_annotation(soft_file)
# 2. Extract gene mapping from annotation data
mapping_data = get_gene_mapping(gene_metadata, prob_col='ID', gene_col='GENE_SYMBOL')
# 3. Apply mapping to convert probe-level data to gene-level data
gene_expression_data = apply_gene_mapping(expression_df=gene_data, mapping_df=mapping_data)
# Save processed gene data
gene_expression_data.to_csv(out_gene_data_file)
# Print shape before and after mapping to verify the transformation
print("Shape before mapping (probes):", gene_data.shape)
print("Shape after mapping (genes):", gene_expression_data.shape)
print("\nFirst few gene symbols and their expression values:")
print(gene_expression_data.head())
# First get clinical features
selected_clinical_df = geo_select_clinical_features(
clinical_df=clinical_data,
trait=trait,
trait_row=0, # From trait identification in previous step
convert_trait=lambda x: 1 if not pd.isna(x) else None, # Initially mark all as patients
age_row=1, # From age identification in previous step
convert_age=convert_age,
gender_row=2, # From gender identification in previous step
convert_gender=convert_gender
)
# Set first 10 samples as controls based on background info
sample_cols = selected_clinical_df.columns[:10] # First 10 samples
selected_clinical_df.loc[trait, sample_cols] = 0
# 1. Normalize gene symbols
gene_expression_data = normalize_gene_symbols_in_index(gene_expression_data)
gene_expression_data.to_csv(out_gene_data_file)
# 2. Link clinical and genetic data
linked_data = geo_link_clinical_genetic_data(selected_clinical_df, gene_expression_data)
# 3. Handle missing values
linked_data = handle_missing_values(linked_data, trait)
# 4. Check for bias
trait_biased, linked_data = judge_and_remove_biased_features(linked_data, trait)
# 5. Validate 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=trait_biased,
df=linked_data,
note="Study examining transcriptome profiles in rheumatoid arthritis."
)
# 6. Save if usable
if is_usable:
linked_data.to_csv(out_data_file)
# Get file paths
soft_file, matrix_file = geo_get_relevant_filepaths(in_cohort_dir)
# Extract gene expression data from matrix file
gene_data = get_genetic_data(matrix_file)
# Print first 20 row IDs and shape of data to help debug
print("Shape of gene expression data:", gene_data.shape)
print("\nFirst few rows of data:")
print(gene_data.head())
print("\nFirst 20 gene/probe identifiers:")
print(gene_data.index[:20])
# Inspect a snippet of raw file to verify identifier format
import gzip
with gzip.open(matrix_file, 'rt', encoding='utf-8') as f:
lines = []
for i, line in enumerate(f):
if "!series_matrix_table_begin" in line:
# Get the next 5 lines after the marker
for _ in range(5):
lines.append(next(f).strip())
break
print("\nFirst few lines after matrix marker in raw file:")
for line in lines:
print(line)
# 1. Extract gene annotation data and observe identifiers
# From previous outputs, we can see:
# - Gene expression data uses identifiers like 'A_19_P00315452'
# - Gene annotation data has matching IDs in the 'ID' column and gene symbols in 'GENE_SYMBOL'
gene_metadata = get_gene_annotation(soft_file)
# 2. Extract gene mapping from annotation data
mapping_data = get_gene_mapping(gene_metadata, prob_col='ID', gene_col='GENE_SYMBOL')
# 3. Apply mapping to convert probe-level data to gene-level data
gene_expression_data = apply_gene_mapping(expression_df=gene_data, mapping_df=mapping_data)
# Save processed gene data
gene_expression_data.to_csv(out_gene_data_file)
# Print shape before and after mapping to verify the transformation
print("Shape before mapping (probes):", gene_data.shape)
print("Shape after mapping (genes):", gene_expression_data.shape)
print("\nFirst few gene symbols and their expression values:")
print(gene_expression_data.head())
# First extract clinical features with proper conversion functions
def convert_trait(x):
if pd.isna(x):
return None
# All samples with tissue:monocytes are trait positive (RA patients) except first 10 which are controls
return 1
def convert_age(x):
if pd.isna(x):
return None
# Extract numeric value after 'age:'
match = re.search(r'age:\s*(\d+)y', str(x))
if match:
return int(match.group(1))
return None
def convert_gender(x):
if pd.isna(x):
return None
# Convert after 'gender:'
if 'female' in str(x).lower():
return 0
elif 'male' in str(x).lower():
return 1
return None
# Extract clinical features
selected_clinical_df = geo_select_clinical_features(
clinical_df=clinical_data,
trait=trait,
trait_row=0, # Using tissue row
convert_trait=convert_trait,
age_row=1, # Age information is in row 1
convert_age=convert_age,
gender_row=2, # Gender information is in row 2
convert_gender=convert_gender
)
# Set first 10 samples as controls based on background info
sample_cols = selected_clinical_df.columns[:10] # First 10 samples
selected_clinical_df.loc[trait, sample_cols] = 0
# 1. Normalize gene symbols from previous gene mapping result
gene_data = normalize_gene_symbols_in_index(gene_data)
gene_data.to_csv(out_gene_data_file)
# 2. Link clinical and genetic data
linked_data = geo_link_clinical_genetic_data(selected_clinical_df, gene_data)
# 3. Handle missing values
linked_data = handle_missing_values(linked_data, trait)
# 4. Check for bias
trait_biased, linked_data = judge_and_remove_biased_features(linked_data, trait)
# 5. Validate 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=trait_biased,
df=linked_data,
note="Study examining transcriptome profiles in rheumatoid arthritis, with 10 controls and 21 RA patients."
)
# 6. Save if usable
if is_usable:
linked_data.to_csv(out_data_file)
# Get file paths
soft_file, matrix_file = geo_get_relevant_filepaths(in_cohort_dir)
# Extract background info and clinical data
background_info, clinical_data = get_background_and_clinical_data(matrix_file)
# Get unique values per clinical feature
sample_characteristics = get_unique_values_by_row(clinical_data)
# Print background info
print("Dataset Background Information:")
print(f"{background_info}\n")
# Print sample characteristics
print("Sample Characteristics:")
for feature, values in sample_characteristics.items():
print(f"Feature: {feature}")
print(f"Values: {values}\n")
# 1. Gene Expression Data Availability
# The series title and summary indicate gene expression data of monocytes
is_gene_available = True
# 2.1 Data Availability
# For trait: While we know there are healthy controls and RA patients from the series design,
# the treatment information is not shown in the available sample characteristics preview
# So we cannot reliably extract trait information
trait_row = None
# Age is in Feature 1
age_row = 1
# Gender is in Feature 2
gender_row = 2
# 2.2 Data Type Conversion Functions
def convert_trait(x):
# Not needed since trait_row is None
return None
def convert_age(x):
if pd.isna(x):
return None
# Extract number before 'y'
try:
age = int(x.split(': ')[1].replace('y',''))
return age
except:
return None
def convert_gender(x):
if pd.isna(x):
return None
val = x.split(': ')[1].lower()
if 'female' in val:
return 0
elif 'male' in val:
return 1
return None
# 3. 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
# Skip since trait_row is None
# Request to see sample characteristics data first
print("Please provide previous output containing:")
print("1. The sample characteristics dictionary")
print("2. Background information about the dataset")
print("3. Any other relevant metadata")
# Set availability flag for gene expression data based on series type
is_gene_available = False # Only miRNA data based on previous output shown
# Define row indices and conversion functions for clinical features
trait_row = None # No disease status/RA information found in sample characteristics
age_row = None # Age information not provided
gender_row = None # Gender information not provided
def convert_trait(x: str) -> int:
return None # Not used since trait_row is None
def convert_age(x: str) -> float:
return None # Not used since age_row is None
def convert_gender(x: str) -> int:
return None # Not used since gender_row is None
# Save initial filtering results
validate_and_save_cohort_info(
is_final=False,
cohort=cohort,
info_path=json_path,
is_gene_available=is_gene_available,
is_trait_available=(trait_row is not None)
)
# Skip clinical feature extraction since trait_row is None