ASCARIS.py

import streamlit as st
import pandas as pd 
from os import path
import sys
import streamlit.components.v1 as components
sys.path.append('code/')
#sys.path.append('ASCARIS/code/') 
import pdb_featureVector
import alphafold_featureVector
import argparse
from st_aggrid import AgGrid, GridOptionsBuilder, JsCode,GridUpdateMode
import base64
showWarningOnDirectExecution = False

def convert_df(df):
   return df.to_csv(index=False, sep='\t').encode('utf-8')

    
# Check if 'key' already exists in session_state
# If not, then initialize it
if 'visibility' not in st.session_state:
    st.session_state['visibility'] = 'visible'
    st.session_state.disabled =  False


showWarningOnDirectExecution = False
original_title = '<p style="font-family:Trebuchet MS; color:#000000; font-size: 25px; font-weight:bold; text-align:center">ASCARIS</p>'
st.markdown(original_title, unsafe_allow_html=True)
original_title = '<p style="font-family:Trebuchet MS; color:#000000; font-size: 25px; font-weight:bold; text-align:center">(Annotation and StruCture-bAsed RepresentatIon of Single amino acid variations)</p>'
st.markdown(original_title, unsafe_allow_html=True)


st.write('')
st.write('')
st.write('')
st.write('')


with st.form('mform', clear_on_submit=False):
    source = st.selectbox('Select the protein structure resource (1: PDB-SwissModel-Modbase, 2: AlphaFold)',[1,2])
    #source = 1
    impute = st.selectbox('Missing value imputation (mostly for the cases where the corresponding annotation does not exist in the protein)',[True, False])
    input_data = st.text_input('Enter SAV data points (format: "UniProt/Swiss-Prot human protein accession" – "wild type a.a." – "position on the sequence" – "mutated a.a.").   Example: Q9BTP7-S-126-F or P04217-A-493-S, Q00889-G-2-L')
            
    parser = argparse.ArgumentParser(description='ASCARIS')

    input_set = input_data
    mode = source
    impute = impute
    submitted = st.form_submit_button(label="Submit", help=None, on_click=None, args=None, kwargs=None, type="secondary", disabled=False, use_container_width=False)
    print('*****************************************')
    print('Feature vector generation is in progress. \nPlease check log file for updates..')
    print('*****************************************')
    mode = int(mode)
    
    
selected_df = pd.DataFrame()
st.write('The online tool may be slow, especially while processing multiple SAVs and with multiple PDB matches. To address this, please consider using the programmatic version at https://github.com/HUBioDataLab/ASCARIS/')

if submitted:
    with st.spinner('In progress...This may take a while...'):
        # try:
        if mode == 1:
            selected_df = pdb_featureVector.pdb(input_set, mode, impute)  
        elif mode == 2:
            selected_df = alphafold_featureVector.alphafold(input_set, mode, impute)
        else:
            selected_df =  pd.DataFrame()

    if selected_df is None:
        st.success('Feature vector failed.')

    else:
        if len(selected_df) != 0 :
            st.write(selected_df)
            st.success('Feature vector successfully created.')
            csv = convert_df(selected_df)
            
    
            st.download_button("Press to Download the Feature Vector", csv,f"ASCARIS_SAV_rep_{input_set}.csv","text/csv",key='download-csv')

        else:
            st.success('Feature vector failed.')