#!/usr/bin/env python3 """ Generate an HTML summary page for a variant that displays all plots in an easily-readable format. Usage: python generate_variant_summary_html.py """ import sys import os import pandas as pd from pathlib import Path from typing import List, Optional, Dict def get_prioritized_models_from_tsv(clustered_tsv: Path, variant_id: str, model_dataset: str) -> List[str]: """Get prioritized models for a variant from clustered.tsv. Reads the models_prioritized_by_any-{model_dataset} column and parses the semicolon-separated format to extract model names. Parameters: ----------- clustered_tsv : Path Path to clustered.tsv file variant_id : str Variant ID to look up model_dataset : str Model dataset name (e.g., "Fetal Brain") Returns: -------- List[str] List of model names that prioritize this variant """ if not clustered_tsv.exists(): print(f"Warning: clustered.tsv does not exist: {clustered_tsv}") return [] try: df = pd.read_csv(clustered_tsv, sep='\t') except Exception as e: print(f"Error reading clustered.tsv: {e}") return [] variant_row = df[df['variant_id'] == variant_id] if variant_row.empty: print(f"Warning: Variant {variant_id} not found in clustered.tsv") return [] models = [] # First, try the aggregated column for this model_dataset aggregated_col = f'models_prioritized_by_any-{model_dataset}' if aggregated_col in df.columns: models_str = variant_row[aggregated_col].iloc[0] if pd.notna(models_str) and str(models_str).strip(): # Parse: ";MODEL1(score);MODEL2(score);..." model_entries = [m.strip() for m in str(models_str).split(';') if m.strip()] for entry in model_entries: if '(' in entry: model = entry.split('(')[0] if model: models.append(model) # If no models found, check all models_prioritized_by_any-* columns # (fallback for different column naming) if not models: for col in df.columns: if col.startswith('models_prioritized_by_any-'): models_str = variant_row[col].iloc[0] if pd.notna(models_str) and str(models_str).strip(): model_entries = [m.strip() for m in str(models_str).split(';') if m.strip()] for entry in model_entries: if '(' in entry: model = entry.split('(')[0] if model: models.append(model) return list(set(models)) # Remove duplicates def find_expected_plots(variant_dir: Path, variant_id: str, clustered_tsv: Path, model_dataset: str) -> dict: """Find all expected plot paths for a variant, regardless of whether they exist. Returns: dict with keys: 'barplot', 'barplot_superset', 'scatterplot', 'scatterplot_superset', 'profiles' (list) """ plots = { 'barplot': variant_dir / "01-model-specificity-barplot.png", 'barplot_superset': variant_dir / "01-model-specificity-barplot-superset.png", 'scatterplot': variant_dir / "02-model-scatterplot.html", 'scatterplot_superset': variant_dir / "02-model-scatterplot-superset.html", 'profiles': [] } # Get prioritized models from clustered.tsv prioritized_models = get_prioritized_models_from_tsv(clustered_tsv, variant_id, model_dataset) for model in prioritized_models: profile_path = variant_dir / f"03-profile-{model}.png" plots['profiles'].append({ 'path': profile_path, 'model': model, 'exists': profile_path.exists() }) return plots def find_plots(variant_dir: Path) -> dict: """Find all plot files in the variant directory. Returns: dict with keys: 'barplot', 'barplot_superset', 'scatterplot', 'scatterplot_superset', 'profiles' (list) """ plots = { 'barplot': None, 'barplot_superset': None, 'scatterplot': None, 'scatterplot_superset': None, 'profiles': [] } # Find cluster-level barplot barplot_path = variant_dir / "01-model-specificity-barplot.png" if barplot_path.exists(): plots['barplot'] = barplot_path # Find superset-level barplot barplot_superset_path = variant_dir / "01-model-specificity-barplot-superset.png" if barplot_superset_path.exists(): plots['barplot_superset'] = barplot_superset_path # Find cluster-level scatterplot scatterplot_path = variant_dir / "02-model-scatterplot.html" if scatterplot_path.exists(): plots['scatterplot'] = scatterplot_path # Find superset-level scatterplot scatterplot_superset_path = variant_dir / "02-model-scatterplot-superset.html" if scatterplot_superset_path.exists(): plots['scatterplot_superset'] = scatterplot_superset_path # Find all profile plots (03-profile-*.png) for profile_file in sorted(variant_dir.glob("03-profile-*.png")): # Extract model name from filename (03-profile-MODEL_NAME.png) model_name = profile_file.stem.replace("03-profile-", "") plots['profiles'].append({ 'path': profile_file, 'model': model_name }) return plots def generate_html(variant_dir: Path, variant_id: str, output_html: Path, plots: dict): """Generate HTML summary page for the variant.""" # Get relative paths for images (relative to output HTML location) output_dir = output_html.parent def get_relative_path(file_path: Path) -> str: """Get relative path from output HTML to the file.""" try: return os.path.relpath(file_path, output_dir) except ValueError: # If on different drives (Windows), use absolute path return str(file_path) html_content = f""" Variant Summary: {variant_id}

Variant Summary

{variant_id}
""" # Add barplot section (cluster-level) html_content += '
\n' html_content += '

Model Specificity Barplot (Cluster-level)

\n' barplot_path = plots.get('barplot') if barplot_path: # Check if file exists (could be Path object or None) if isinstance(barplot_path, Path) and barplot_path.exists(): rel_path = get_relative_path(barplot_path) html_content += f'''
Model prioritization across tissues/organs in the current cluster/model dataset
Model Specificity Barplot
''' else: # File doesn't exist yet, but include it anyway (browser will show broken link until file appears) rel_path = get_relative_path(barplot_path) html_content += f'''
Model prioritization across tissues/organs in the current cluster/model dataset
Model Specificity Barplot
Plot will be available after generation
''' else: html_content += '
No cluster-level barplot available
\n' html_content += '
\n\n' # Add superset barplot section html_content += '
\n' html_content += '

Model Specificity Barplot (Superset-level)

\n' barplot_superset_path = plots.get('barplot_superset') if barplot_superset_path: if isinstance(barplot_superset_path, Path) and barplot_superset_path.exists(): rel_path = get_relative_path(barplot_superset_path) html_content += f'''
Model prioritization across tissues/organs in the superset (broader context)
Model Specificity Barplot (Superset)
''' else: rel_path = get_relative_path(barplot_superset_path) html_content += f'''
Model prioritization across tissues/organs in the superset (broader context)
Model Specificity Barplot (Superset)
Plot will be available after generation
''' else: html_content += '
No superset-level barplot available
\n' html_content += '
\n\n' # Add scatterplot section (cluster-level) html_content += '
\n' html_content += '

Model Scatterplot (Cluster-level)

\n' scatterplot_path = plots.get('scatterplot') if scatterplot_path: if isinstance(scatterplot_path, Path) and scatterplot_path.exists(): rel_path = get_relative_path(scatterplot_path) html_content += f'''
Interactive scatterplot showing variant scores across models in the current cluster/model dataset
''' else: rel_path = get_relative_path(scatterplot_path) html_content += f'''
Interactive scatterplot showing variant scores across models in the current cluster/model dataset
Plot will be available after generation
''' else: html_content += '
No cluster-level scatterplot available
\n' html_content += '
\n\n' # Add superset scatterplot section html_content += '
\n' html_content += '

Model Scatterplot (Superset-level)

\n' scatterplot_superset_path = plots.get('scatterplot_superset') if scatterplot_superset_path: if isinstance(scatterplot_superset_path, Path) and scatterplot_superset_path.exists(): rel_path = get_relative_path(scatterplot_superset_path) html_content += f'''
Interactive scatterplot showing variant scores across all models in the superset (broader context)
''' else: rel_path = get_relative_path(scatterplot_superset_path) html_content += f'''
Interactive scatterplot showing variant scores across all models in the superset (broader context)
Plot will be available after generation
''' else: html_content += '
No superset-level scatterplot available
\n' html_content += '
\n\n' # Add profiles section html_content += '
\n' html_content += '

Profile Plots

\n' if plots['profiles']: html_content += '
\n' for profile in plots['profiles']: profile_path = profile['path'] rel_path = get_relative_path(profile_path) exists = profile.get('exists', False) if exists: html_content += f'''

{profile['model']}

Profile plot for {profile['model']}
''' else: html_content += f'''

{profile['model']}

Profile plot for {profile['model']}
Plot will be available after generation
''' html_content += '
\n' else: html_content += '
No profile plots available
\n' html_content += '
\n\n' html_content += """ """ # Write HTML file output_html.parent.mkdir(parents=True, exist_ok=True) with open(output_html, 'w') as f: f.write(html_content) print(f"✓ Generated HTML summary: {output_html}") def main(): if len(sys.argv) != 6: print("Usage: python generate_variant_summary_html.py ") sys.exit(1) variant_dir = Path(sys.argv[1]) variant_id = sys.argv[2] clustered_tsv = Path(sys.argv[3]) model_dataset = sys.argv[4] output_html = Path(sys.argv[5]) # Create variant directory if it doesn't exist (plots may not be generated yet) variant_dir.mkdir(parents=True, exist_ok=True) # Find all expected plots (even if they don't exist yet) plots = find_expected_plots(variant_dir, variant_id, clustered_tsv, model_dataset) # Generate HTML generate_html(variant_dir, variant_id, output_html, plots) # Print summary num_profiles = len(plots['profiles']) print(f"Summary: Expected {num_profiles} profile plot(s)") for profile in plots['profiles']: status = "✓" if profile.get('exists', False) else "(pending)" print(f" - {profile['model']}: {status}") barplot_exists = isinstance(plots.get('barplot'), Path) and plots['barplot'].exists() barplot_superset_exists = isinstance(plots.get('barplot_superset'), Path) and plots['barplot_superset'].exists() scatterplot_exists = isinstance(plots.get('scatterplot'), Path) and plots['scatterplot'].exists() scatterplot_superset_exists = isinstance(plots.get('scatterplot_superset'), Path) and plots['scatterplot_superset'].exists() print(f" - Barplot: {'✓' if barplot_exists else '(pending)'}") print(f" - Barplot (superset): {'✓' if barplot_superset_exists else '(pending)'}") print(f" - Scatterplot: {'✓' if scatterplot_exists else '(pending)'}") print(f" - Scatterplot (superset): {'✓' if scatterplot_superset_exists else '(pending)'}") if __name__ == "__main__": main()