prevent crash when off target is empty and add option for on target only

app.py

@@ -15,6 +15,11 @@ st.session_state["fasta_seq"] = ""
 st.session_state["text_seq"] = ""
 status_bar, status_text = None, None
 
+with st.form(key='calc_options'):
+    c1_names = ['On Target Only', 'On and Off Target']
+    option = st.radio('Calculation Options', c1_names, index=0)
+    submitButton = st.form_submit_button(label='Choose Options')
+
 # UserInput Form from text input
 text_form = st.form("text")
 text_input = text_form.text_input(
@@ -73,7 +78,7 @@ else:
 # valid input
 if src_seq and all([True if nt.upper() in NUCLEOTIDE_TOKENS.keys() else False for nt in src_seq]):
     on_target, off_target = tiger_exhibit(pd.DataFrame(dict(id=['ManualEntry'], seq=[src_seq])),
-            status_bar, status_text)
+            status_bar, status_text, option)
     st.write('On-target predictions: ', on_target)
     st.download_button(label='Download', data=convert_df(on_target), file_name='on_target.csv', mime='text/csv')
     if len(off_target) > 0:

tiger.py

@@ -188,7 +188,7 @@ def predict_off_target(off_targets: pd.DataFrame, model: tf.keras.Model):
     return off_targets.sort_values('Normalized LFC')
 
 
-def tiger_exhibit(transcripts: pd.DataFrame, status_bar=None, status_text=None):
+def tiger_exhibit(transcripts: pd.DataFrame, status_bar=None, status_text=None, option=''):
 
     # load model
     if os.path.exists('model'):
@@ -213,8 +213,15 @@ def tiger_exhibit(transcripts: pd.DataFrame, status_bar=None, status_text=None):
     print('')
 
     # predict off-target effects for top guides
-    off_targets = find_off_targets(on_target_predictions,  status_bar, status_text)
-    off_target_predictions = predict_off_target(off_targets, model=tiger)
+    off_target_predictions = pd.DataFrame()
+    if option=='On and Off Target':
+        off_targets = find_off_targets(on_target_predictions,  status_bar, status_text)
+        off_target_predictions = predict_off_target(off_targets, model=tiger)
+
+    # reverse guide sequences
+    on_target_predictions['Guide'] = on_target_predictions['Guide'].apply(lambda s: s[::-1])
+    if option=='On and Off Target' and len(off_target_predictions) > 0:
+        off_target_predictions['Guide'] = off_target_predictions['Guide'].apply(lambda s: s[::-1])
 
     return on_target_predictions.reset_index(drop=True), off_target_predictions.reset_index(drop=True)