Update app.py

app.py

@@ -46,35 +46,31 @@ selected_features = [
 ]
 
 def extract_features(sequence):
-    if len(sequence) < 2:  # Ensure sequence is long enough
-        return None  
+    if len(sequence) < 3:  # Ensure sequence is long enough
+        return None  # Return None if sequence is too short
+
+    dipeptide_features = AAComposition.CalculateAADipeptideComposition(sequence)
+    auto_features = Autocorrelation.CalculateAutoTotal(sequence)
+    ctd_features = CTD.CalculateCTD(sequence)
 
     try:
-        dipeptide_features = AAComposition.CalculateAADipeptideComposition(sequence)
-        auto_features = Autocorrelation.CalculateAutoTotal(sequence)
-        ctd_features = CTD.CalculateCTD(sequence)
-        
-        # Check sequence length before APAAC
-        if len(sequence) > 2:  # APAAC requires a longer sequence
-            pseudo_features = PseudoAAC.GetAPseudoAAC(sequence)
-        else:
-            pseudo_features = {}
-
-        all_features = {**auto_features, **ctd_features, **pseudo_features, **dipeptide_features}
-        all_features = list(all_features.values())
-        all_features = np.array(all_features).reshape(-1, 1)  # Correct shape
-        normalized_features = scaler.transform(all_features.T)
-        normalized_features = normalized_features.flatten()
-
-        selected_feature_dict = {feature: normalized_features[i] for i, feature in enumerate(selected_features) if feature in all_features}
-        selected_feature_df = pd.DataFrame([selected_feature_dict])
-        selected_feature_array = selected_feature_df.T.to_numpy()
-
-        return selected_feature_array
-
-    except Exception as e:
-        print(f"Feature extraction error: {e}")
-        return None  # Return None if extraction fails
+        pseudo_features = PseudoAAC.GetAPseudoAAC(sequence)  # May cause ZeroDivisionError
+    except ZeroDivisionError:
+        pseudo_features = {}  # Ignore if it fails
+
+    all_features = {**auto_features, **ctd_features, **pseudo_features, **dipeptide_features}
+    all_features = list(all_features.values())
+    all_features = np.array(all_features).reshape(-1, 1)  # Correct shape
+
+    normalized_features = scaler.transform(all_features.T)
+    normalized_features = normalized_features.flatten()  
+
+    selected_feature_dict = {feature: normalized_features[i] for i, feature in enumerate(selected_features) if feature in all_features}
+    selected_feature_df = pd.DataFrame([selected_feature_dict])
+    selected_feature_array = selected_feature_df.T.to_numpy()
+
+    return selected_feature_array
+