Add debug logging to track DNA optimization flow

- Add logging for protein translation steps
- Track sequence lengths and content through optimization
- Debug why DNA input produces identical output sequences
- This will help identify if issue is in translation, model, or comparison

app.py

@@ -595,6 +595,12 @@ def run_optimization(protein: str, organism: str, use_post_processing: bool = Fa
     st.session_state.optimization_running = True
     st.session_state.post_processed_results = None
 
+    # Debug logging
+    print(f"🔍 DEBUG: Starting optimization")
+    print(f"🔍 DEBUG: Protein length: {len(protein)}")
+    print(f"🔍 DEBUG: Protein (first 50): {protein[:50]}...")
+    print(f"🔍 DEBUG: Organism: {organism}")
+
     try:
         # Use the exact same method that achieved best results in evaluation
         result = predict_dna_sequence(
@@ -606,6 +612,11 @@ def run_optimization(protein: str, organism: str, use_post_processing: bool = Fa
             match_protein=True,
         )
 
+        # Debug logging for result
+        _res = result[0] if isinstance(result, list) else result
+        print(f"🔍 DEBUG: Model returned DNA length: {len(_res.predicted_dna)}")
+        print(f"🔍 DEBUG: Model returned DNA (first 50): {_res.predicted_dna[:50]}...")
+
         # Check GC content and auto-correct if out of optimal range
         _res = result[0] if isinstance(result, list) else result
         initial_gc = get_GC_content(_res.predicted_dna)
@@ -867,7 +878,12 @@ def single_sequence_optimization():
             if st.button("🚀 Optimize Sequence", type="primary", use_container_width=True):
                 st.session_state.results = None
                 if st.session_state.sequence_type == "dna":
-                    protein_sequence = translate_dna_to_protein(str(st.session_state.sequence_clean))
+                    original_dna = str(st.session_state.sequence_clean)
+                    protein_sequence = translate_dna_to_protein(original_dna)
+                    print(f"🔍 DEBUG: Original DNA length: {len(original_dna)}")
+                    print(f"🔍 DEBUG: Translated protein length: {len(protein_sequence)}")
+                    print(f"🔍 DEBUG: Original DNA (first 50): {original_dna[:50]}...")
+                    print(f"🔍 DEBUG: Translated protein (first 30): {protein_sequence[:30]}...")
                     run_optimization(protein_sequence, str(st.session_state.organism), use_post_processing)
                 else:
                     run_optimization(str(st.session_state.sequence_clean), str(st.session_state.organism), use_post_processing)
@@ -950,11 +966,16 @@ def display_optimization_results(result, organism, original_sequence, sequence_t
         with imp_col1:
             if input_metrics.get('gc_content') and optimized_metrics.get('gc_content'):
                 gc_change = optimized_metrics['gc_content'] - input_metrics['gc_content']
+                print(f"🔍 DEBUG: GC change: {input_metrics['gc_content']:.3f} -> {optimized_metrics['gc_content']:.3f} (Δ{gc_change:+.3f})")
                 st.metric("GC Content", f"{optimized_metrics['gc_content']:.1f}%", delta=f"{gc_change:+.1f}%")
 
         with imp_col2:
             if input_metrics.get('cai') and optimized_metrics.get('cai'):
                 cai_change = optimized_metrics['cai'] - input_metrics['cai']
+                print(f"🔍 DEBUG: CAI change: {input_metrics['cai']:.6f} -> {optimized_metrics['cai']:.6f} (Δ{cai_change:+.6f})")
+                print(f"🔍 DEBUG: Are sequences identical? {original_sequence == result.predicted_dna}")
+                if hasattr(result, 'predicted_dna') and len(original_sequence) > 0:
+                    print(f"🔍 DEBUG: Sequence lengths - Original: {len(original_sequence)}, Optimized: {len(result.predicted_dna)}")
                 st.metric("CAI Score", f"{optimized_metrics['cai']:.3f}", delta=f"{cai_change:+.3f}")
 
         with imp_col3: