Update app.py

app.py

@@ -204,51 +204,91 @@ def run_pipeline(dna_input):
                         logging.info("F gene sequence found in dataset")
                         matched_ids, perc = analyzer.find_similar_sequences(analyzer.matching_percentage)
                         
-                        # Try to create tree with error handling
+                        # Enhanced tree creation with better error handling
                         try:
-                            analyzer.create_interactive_tree(matched_ids, perc)
+                            logging.info(f"Creating tree with {len(matched_ids)} sequences at {perc:.1f}% similarity")
                             
-                            # Check for multiple possible HTML file names
+                            # Call the tree creation method
+                            tree_result = analyzer.create_interactive_tree(matched_ids, perc)
+                            
+                            # Check for multiple possible HTML file names and locations
                             possible_html_files = [
                                 "phylogenetic_tree_normalized_horizontal.html",
+                                "phylogenetic_tree_horizontal.html", 
                                 "phylogenetic_tree.html",
-                                "tree.html"
+                                "tree.html",
+                                "interactive_tree.html"
                             ]
                             
-                            for html_filename in possible_html_files:
-                                if os.path.exists(html_filename):
-                                    html_file = html_filename
-                                    with open(html_filename, "r", encoding='utf-8') as f:
-                                        tree_html_content = f.read()
-                                    ml_output = f"Phylogenetic tree generated successfully with {len(matched_ids)} sequences (similarity: {perc:.1f}%)"
-                                    logging.info(f"Tree HTML file found: {html_filename}")
+                            # Also check in common subdirectories
+                            search_dirs = [".", "output", "results", "trees"]
+                            
+                            for search_dir in search_dirs:
+                                if not os.path.exists(search_dir):
+                                    continue
+                                    
+                                for html_filename in possible_html_files:
+                                    full_path = os.path.join(search_dir, html_filename)
+                                    if os.path.exists(full_path):
+                                        html_file = full_path
+                                        try:
+                                            with open(full_path, "r", encoding='utf-8') as f:
+                                                tree_html_content = f.read()
+                                            ml_output = f"✅ Phylogenetic tree generated successfully!\n- {len(matched_ids)} sequences analyzed\n- Similarity threshold: {perc:.1f}%\n- Tree file: {html_filename}"
+                                            logging.info(f"Tree HTML file found: {full_path}")
+                                            break
+                                        except Exception as read_error:
+                                            logging.error(f"Error reading HTML file {full_path}: {read_error}")
+                                            continue
+                                
+                                if html_file:
                                     break
                             
                             if not html_file:
-                                # List all HTML files in directory for debugging
-                                html_files = [f for f in os.listdir('.') if f.endswith('.html')]
-                                ml_output = f"Tree analysis completed but HTML file not found. Available HTML files: {html_files}"
-                                logging.warning(f"HTML files in directory: {html_files}")
+                                # Enhanced debugging - list all files in current directory
+                                all_files = []
+                                for root, dirs, files in os.walk("."):
+                                    for file in files:
+                                        if file.endswith(('.html', '.htm')):
+                                            all_files.append(os.path.join(root, file))
+                                
+                                ml_output = f"⚠️ Tree analysis completed but HTML file not found.\n"
+                                ml_output += f"- Analyzed {len(matched_ids)} sequences at {perc:.1f}% similarity\n"
+                                ml_output += f"- Available HTML files in directory: {all_files}\n"
+                                ml_output += f"- Tree creation result: {tree_result if 'tree_result' in locals() else 'Unknown'}"
+                                
+                                logging.warning(f"HTML files found in directory tree: {all_files}")
+                                
+                                # Try to get any HTML content that might have been generated
+                                if hasattr(analyzer, 'last_generated_html') and analyzer.last_generated_html:
+                                    tree_html_content = analyzer.last_generated_html
+                                    ml_output += "\n- Using HTML content from analyzer object"
                         
                         except Exception as tree_error:
-                            ml_output = f"Tree creation failed: {str(tree_error)}"
+                            ml_output = f"❌ Tree creation failed: {str(tree_error)}\n"
+                            ml_output += f"- Sequences available: {len(matched_ids)}\n"
+                            ml_output += f"- Similarity threshold: {perc:.1f}%"
                             logging.error(f"Tree creation error: {tree_error}")
+                            import traceback
+                            logging.error(f"Full traceback: {traceback.format_exc()}")
                             
                     else:
-                        ml_output = f"F gene sequence not found in dataset (length: {len(processed_sequence)} bp)"
+                        ml_output = f"❌ F gene sequence not found in dataset\n- Query length: {len(processed_sequence)} bp\n- Dataset contains {len(analyzer.data) if hasattr(analyzer, 'data') else 'unknown'} sequences"
                         logging.warning(f"F gene sequence not found. Length: {len(processed_sequence)}")
                 else:
-                    ml_output = "Failed to load CSV dataset"
+                    ml_output = "❌ Failed to load CSV dataset"
                     logging.error("Failed to load CSV dataset")
             except Exception as e:
-                ml_output = f"ML Tree analysis failed: {str(e)}"
+                ml_output = f"❌ ML Tree analysis failed: {str(e)}"
                 logging.error(f"ML Tree failed: {e}")
+                import traceback
+                logging.error(f"Full traceback: {traceback.format_exc()}")
         elif not os.path.exists(csv_path):
-            ml_output = f"CSV dataset not found at {csv_path}"
+            ml_output = f"❌ CSV dataset not found at {csv_path}"
         elif not processed_sequence or len(processed_sequence) < 10:
-            ml_output = f"F gene sequence too short for analysis (length: {len(processed_sequence) if processed_sequence else 0})"
+            ml_output = f"❌ F gene sequence too short for analysis (length: {len(processed_sequence) if processed_sequence else 0})"
         else:
-            ml_output = "Skipped due to previous step errors"
+            ml_output = "❌ Skipped due to previous step errors"
 
         return (
             boundary_output,
@@ -264,45 +304,48 @@ def run_pipeline(dna_input):
     except Exception as e:
         error_msg = f"Pipeline failed: {str(e)}"
         logging.error(error_msg)
+        import traceback
+        logging.error(f"Full traceback: {traceback.format_exc()}")
         return error_msg, "", "", "", None, None, None, error_msg
 
 # --- Gradio UI ---
-with gr.Blocks(title="Viral Gene Phylogenetic Pipeline") as demo:
-    gr.Markdown("# Viral Gene Phylogenetic Inference Pipeline")
+with gr.Blocks(title="Viral Gene Phylogenetic Pipeline", theme=gr.themes.Soft()) as demo:
+    gr.Markdown("# 🧬 Viral Gene Phylogenetic Inference Pipeline")
     gr.Markdown("This pipeline processes DNA sequences through boundary detection, k-mer analysis, and phylogenetic tree construction.")
 
-    with gr.Tab("Paste DNA Sequence"):
+    with gr.Tab("📝 Paste DNA Sequence"):
         inp = gr.Textbox(
             label="DNA Input", 
             placeholder="Paste your DNA sequence here (ACTG format)",
             lines=5
         )
-        btn1 = gr.Button("Run Pipeline", variant="primary")
+        btn1 = gr.Button("🚀 Run Pipeline", variant="primary", size="lg")
 
-    with gr.Tab("Upload FASTA File"):
+    with gr.Tab("📁 Upload FASTA File"):
         file_input = gr.File(
             label="FASTA File", 
             file_types=['.fasta', '.fa', '.txt']
         )
-        btn2 = gr.Button("Run on FASTA", variant="primary")
+        btn2 = gr.Button("🚀 Run on FASTA", variant="primary", size="lg")
 
     # Outputs
-    gr.Markdown("## Pipeline Results")
+    gr.Markdown("## 📊 Pipeline Results")
     
     with gr.Row():
         with gr.Column():
-            out1 = gr.Textbox(label="Step 1: Extracted F Gene Sequence", lines=8)
-            out2 = gr.Textbox(label="Step 2: F Gene Validation (Keras)", lines=3)
+            out1 = gr.Textbox(label="🎯 Step 1: Extracted F Gene Sequence", lines=8)
+            out2 = gr.Textbox(label="🔍 Step 2: F Gene Validation (Keras)", lines=3)
         with gr.Column():
-            out3 = gr.Textbox(label="Dataset Used")
-            out4 = gr.Textbox(label="Step 3: Phylogenetic Tree Status", lines=3)
+            out3 = gr.Textbox(label="📋 Dataset Used")
+            out4 = gr.Textbox(label="🌳 Step 3: Phylogenetic Tree Status", lines=5)
 
     with gr.Row():
-        html = gr.File(label="Download Tree (HTML)")
-        fasta = gr.File(label="Download Aligned FASTA")
-        phy = gr.File(label="Download IQ-TREE .phy File")
+        html = gr.File(label="📥 Download Tree (HTML)")
+        fasta = gr.File(label="📥 Download Aligned FASTA")
+        phy = gr.File(label="📥 Download IQ-TREE .phy File")
 
-    tree_html = gr.HTML(label="Interactive Tree Preview")
+    with gr.Row():
+        tree_html = gr.HTML(label="🌳 Interactive Tree Preview", height=600)
 
     # Event handlers
     btn1.click(