results on one page

app.py

@@ -22,15 +22,10 @@ st.markdown("""
 
 
 
-domain_tab, pred_tab = st.tabs(['Domains', 'Function predictions'])
-
-with domain_tab:
-    st.header('Domains in sequence')
-
 with st.sidebar:
 
     st.title("Domain2GO: Mutual Annotation-Based Prediction of Protein Domain Functions")
-    st.write("[![arXiv](https://img.shields.io/badge/bioRxiv-2022.11.03.514980-b31b1b.svg)](https://www.biorxiv.org/content/10.1101/2022.11.03.514980v1) [![github-repository](https://img.shields.io/badge/GitHub-black?logo=github)](https://github.com/HUBioDataLab/Domain2GO)")
+    st.write("[![biorxiv](https://img.shields.io/badge/bioRxiv-2022.11.03.514980-b31b1b.svg)](https://www.biorxiv.org/content/10.1101/2022.11.03.514980v1) [![github-repository](https://img.shields.io/badge/GitHub-black?logo=github)](https://github.com/HUBioDataLab/Domain2GO)")
 
     if 'example_seq_button' not in st.session_state:
         st.session_state.example_seq_button = False
@@ -58,79 +53,81 @@ with st.sidebar:
 
     st.session_state['email'] = st.text_input('Enter your email for InterProScan query: ')
 
-    # prevent user from clicking 'Find domains' button if email or sequence is empty
-    domains_submitted = False
-    if st.button('Find domains'):
+# prevent user from clicking submit button if email or sequence is empty
+submitted = False
+
+
+with st.sidebar:
+    if st.button('Predict functions'):
         if 'email' in st.session_state and 'sequence' in st.session_state and '@' in st.session_state.email:
-            domains_submitted = True
+            submitted = True
             st.session_state.disabled = True
         else:
-            st.warning('Please enter your email and protein sequence first. If you have already entered your email and protein sequence, please check that your email is valid.')
+            with st.sidebar:
+                st.warning('Please enter your email and protein sequence first. If you have already entered your email and protein sequence, please check that your email is valid.')
+
+if not submitted:
+    # on main page, write warning message if user has not submitted email and sequence
+    st.markdown("""
+    <div style="padding:30px">
+    <p style="color:#2a7b36;font-size:20px;">Submit your protein sequence to start.</p>
+    </div>
+    """, unsafe_allow_html=True)
+
+no_domains = False
+error_in_interproscan = False
+if submitted:
+    with st.spinner('Finding domains in sequence using InterProScan. This may take a while...'):
+        result = find_domains(st.session_state.email, st.session_state.sequence, st.session_state.name)
+    result_text = result[0]
+    if result_text == 'Domains found.':
+        # st.success(result_text + ' You can now see function predictions for the sequence in the "Function predictions" tab.')
+        st.session_state['domain_df'] = result[1]
+    elif result_text == 'No domains found.':
+        st.warning(result_text)
+        no_domains = True
     else:
-        with domain_tab:
-            st.warning('Please enter your query and click "Find domains" to see domains in sequence.')
-
-with domain_tab:
-    no_domains = False
-    error_in_interproscan = False
-    if domains_submitted:
-        with st.spinner('Finding domains in sequence using InterProScan. This may take a while...'):
-            result = find_domains(st.session_state.email, st.session_state.sequence, st.session_state.name)
-        result_text = result[0]
-        if result_text == 'Domains found.':
-            st.success(result_text + ' You can now see function predictions for the sequence in the "Function predictions" tab.')
-            st.session_state['domain_df'] = result[1]
-        elif result_text == 'No domains found.':
-            st.warning(result_text)
-            no_domains = True
-        else:
-            st.error(result_text)
-            st.write(f'InterProScan job id: {result[1]}')
-            st.write(f'InterProScan job response: {result[2]}')
-            error_in_interproscan = True
-
-
-    if 'domain_df' in st.session_state:
-        with st.expander('Show domains in sequence'):
-            st.write(st.session_state.domain_df)
-            domains_csv = convert_df(st.session_state.domain_df)
+        st.error(result_text)
+        st.write(f'InterProScan job id: {result[1]}')
+        st.write(f'InterProScan job response: {result[2]}')
+        error_in_interproscan = True
+
+
+# if 'domain_df' in st.session_state:
+#     with st.expander('Show domains in sequence'):
+#         st.write(st.session_state.domain_df)
+#         domains_csv = convert_df(st.session_state.domain_df)
+#         st.download_button(
+#             label="Download domains in sequence as CSV",
+#             data=domains_csv,
+#             file_name=f"{st.session_state.name}_domains.csv",
+#             mime="text/csv",
+#         )
+
+if 'domain_df' not in st.session_state:
+    if error_in_interproscan:
+        st.error('Error in InterProScan. Please check InterProScan job id and response.')
+else:
+    with st.spinner('Generating function predictions...'):
+        cwd = os.getcwd()
+        # mapping_path = "{}/Domain2GO/data".format(cwd.split("Domain2GO")[0])
+        mapping_path = './data'
+        pred_results = generate_function_predictions(st.session_state.domain_df, mapping_path)
+        pred_result_text = pred_results[0]
+        if pred_result_text == 'Function predictions found.':
+            st.success('Function predictions generated.')
+            st.session_state['pred_df'] = pred_results[1]
+        elif pred_result_text == 'No predictions made for domains found in sequence.':
+            st.warning(pred_result_text)
+
+    if 'pred_df' in st.session_state:
+        with st.expander('Show function predictions'):
+            st.write(st.session_state.pred_df)
+            pred_csv = convert_df(st.session_state.pred_df)
             st.download_button(
-                label="Download domains in sequence as CSV",
-                data=domains_csv,
-                file_name=f"{st.session_state.name}_domains.csv",
+                label="Download function predictions as CSV",
+                data=pred_csv,
+                file_name=f"{st.session_state.name}_function_predictions.csv",
                 mime="text/csv",
             )
 
-with pred_tab:
-    st.header('Function predictions')
-    if 'domain_df' not in st.session_state:
-        if no_domains:
-            st.warning('No domains found. Please find domains in sequence first.')
-        elif error_in_interproscan:
-            st.error('Error in InterProScan. Please check InterProScan job id and response.')
-        else:
-            st.warning('Please find domains in sequence first.')
-    else:
-        with st.spinner('Generating function predictions...'):
-            cwd = os.getcwd()
-            # mapping_path = "{}/Domain2GO/data".format(cwd.split("Domain2GO")[0])
-            mapping_path = './data'
-            pred_results = generate_function_predictions(st.session_state.domain_df, mapping_path)
-            pred_result_text = pred_results[0]
-            if pred_result_text == 'Function predictions found.':
-                st.success(pred_result_text)
-                st.session_state['pred_df'] = pred_results[1]
-            elif pred_result_text == 'No function predictions found.':
-                st.warning(pred_result_text)
-
-        if 'pred_df' in st.session_state:
-            with st.expander('Show function predictions'):
-                st.write(st.session_state.pred_df)
-                pred_csv = convert_df(st.session_state.pred_df)
-                st.download_button(
-                    label="Download function predictions as CSV",
-                    data=pred_csv,
-                    file_name=f"{st.session_state.name}_function_predictions.csv",
-                    mime="text/csv",
-                )
-

run_domain2go_app.py

@@ -98,7 +98,7 @@ def generate_function_predictions(domains_df, mapping_path):
     domain2go_df = pd.read_csv(os.path.join(mapping_path, 'finalized_domain2go_mappings.txt'))
     print('Domain2GO mappings loaded')
     # merge domain2go mappings with domains found in protein sequence
-    merged_df = pd.merge(domains_df, domain2go_df, left_on='accession', right_on='Interpro')
+    merged_df = pd.merge(domains_df, domain2go_df, left_on='domain_accession', right_on='Interpro')
 
     print('Function predictions generated.')
     
@@ -109,7 +109,7 @@ def generate_function_predictions(domains_df, mapping_path):
     
     else:
         merged_df['protein_name'] = domains_df['protein_name'].iloc[0]
-        merged_df = merged_df[['protein_name', 'GO', 'locations', 's', 'accession', 'name',]]
+        merged_df = merged_df[['protein_name', 'GO', 'domain_locations', 's', 'domain_accession', 'domain_name',]]
         merged_df.columns = ['protein_name', 'GO_ID', 'domain_locations', 'probability', 'domain_accession', 'domain_name',]
 
         # save protein function predictions