Update app.py

app.py

@@ -46,20 +46,15 @@ selected_features = [
 ]
 
 def extract_features(sequence):
-    """Extract selected features, ensure order matches trained features, and normalize them."""
-    if len(sequence) <= 9:
-        return "Error: Protein sequence must be longer than 9 amino acids to extract features (for lamda=9)."
 
     all_features_dict = {}
 
     # Calculate all dipeptide features
     dipeptide_features = AAComposition.CalculateAADipeptideComposition(sequence)
-    
-    # Add only the first 420 features to the dictionary
-    first_420_keys = list(dipeptide_features.keys())[:420]  # Get the first 420 keys
-    filtered_dipeptide_features = {key: dipeptide_features[key] for key in first_420_keys}
-    
-    all_features_dict.update(filtered_dipeptide_features)
+
+    # Add all dipeptide features
+    all_features_dict.update(dipeptide_features)
+
 
     auto_features = Autocorrelation.CalculateAutoTotal(sequence)
     all_features_dict.update(auto_features)
@@ -70,20 +65,23 @@ def extract_features(sequence):
     pseudo_features = PseudoAAC.GetAPseudoAAC(sequence, lamda=9)
     all_features_dict.update(pseudo_features)
 
-    feature_values = list(all_features_dict.values())
-    feature_array = np.array(feature_values).reshape(-1, 1) # Reshape to (1, n_features) - CORRECT SHAPE
-    print(f"Shape of feature_array before normalization: {feature_array.shape}") # Debug print
+    # Convert feature dictionary to DataFrame, handling missing features
+    feature_df = pd.DataFrame([all_features_dict])
+
+    # Select features and handle missing columns
+    feature_df_selected = feature_df[selected_features].copy() # Use .copy() to avoid SettingWithCopyWarning
+
+    # Fill missing features with 0 (or another appropriate value)
+    feature_df_selected = feature_df_selected.fillna(0)
 
-    normalized_features = scaler.transform(feature_array.T) # Normalize - NO TRANSPOSE
-    normalized_features = normalized_features.flatten() # Flatten AFTER normalization if needed
 
+    feature_array = feature_df_selected.values # Get numpy array directly
 
-    selected_feature_dict = {feature: normalized_features[i] for i, feature in enumerate(selected_features) if feature in all_features_dict}
 
-    selected_feature_df = pd.DataFrame([selected_feature_dict])
-    selected_feature_array = selected_feature_df.T.to_numpy()
+    # Normalize the features
+    normalized_features = scaler.transform(feature_array)
 
-    return selected_feature_array
+    return normalized_features
 
 
 def predict(sequence):
@@ -92,8 +90,8 @@ def predict(sequence):
     if isinstance(features, str) and features.startswith("Error:"):
         return features
 
-    prediction = model.predict(features.T)[0]
-    probabilities = model.predict_proba(features.T)[0]
+    prediction = model.predict(features)[0]
+    probabilities = model.predict_proba(features)[0]
 
     if prediction == 0:
         return f"{probabilities[0] * 100:.2f}% chance of being an Antimicrobial Peptide (AMP)"