app.py

import streamlit as st
import torch
import os
from rdkit import Chem
from rdkit.Chem import Draw
from transformers import BartForConditionalGeneration, BartTokenizer
from admet_ai import ADMETModel
import safe
import io
from PIL import Image
import cairosvg
import pandas as pd

# Page Configuration
st.set_page_config(page_title='Beta-Lactam Molecule Generator', layout='wide')

# Load Models
@st.cache_resource(show_spinner="Loading Models...", ttl=600)
def load_models():
    # Load your molecule generation model
    model_name = "bcadkins01/beta_lactam_generator"  # Replace with your actual model path
    access_token = os.getenv("HUGGING_FACE_TOKEN")
    model = BartForConditionalGeneration.from_pretrained(model_name, use_auth_token=access_token)
    tokenizer = BartTokenizer.from_pretrained(model_name, use_auth_token=access_token)
    # Load ADMET-AI model
    admet_model = ADMETModel()
    return model, tokenizer, admet_model

model, tokenizer, admet_model = load_models()

# Set Generation Parameters
st.sidebar.header('Generation Parameters')
creativity = st.sidebar.slider('Creativity (Temperature):', 0.0, 2.0, 1.0, step=0.1)
num_molecules = st.sidebar.number_input('Number of Molecules to Generate:', min_value=1, max_value=5, value=5)

# String Format Option
string_format = st.sidebar.radio('String Format:', ('SMILES', 'SAFE'))

# Generate Molecules Button
if st.button('Generate Molecules'):
    st.info("Generating molecules... Please wait.")
    # Generate molecules
    core_smiles = "C1C(=O)N(C)C(=O)C1"  # Beta-lactam core structure
    output_ids = model.generate(
        tokenizer(core_smiles, return_tensors='pt').input_ids,
        max_length=128,
        temperature=creativity,
        do_sample=True,
        top_k=50,
        num_return_sequences=num_molecules
    )
    generated_smiles = [tokenizer.decode(ids, skip_special_tokens=True) for ids in output_ids]
    molecule_names = [f"Mol{str(i).zfill(2)}" for i in range(1, num_molecules + 1)]
    generated_molecules = dict(zip(molecule_names, generated_smiles))
    
    # ADMET Predictions
    preds = admet_model.predict(smiles=list(generated_molecules.values()))
    preds['Molecule Name'] = molecule_names
    preds.set_index('Molecule Name', inplace=True)
    
    # Display Molecules
    st.subheader('Generated Molecules')
    cols_per_row = min(5, num_molecules)
    cols = st.columns(cols_per_row)
    for idx, mol_name in enumerate(molecule_names):
        smiles = generated_molecules[mol_name]
        img = generate_molecule_image(smiles, use_safe_visualization=(string_format == 'SAFE'))
        with cols[idx % cols_per_row]:
            if isinstance(img, Image.Image):
                st.image(img, caption=mol_name)
            else:
                st.error(f"Could not generate image for {mol_name}")
            # Display molecule string
            string_to_display = safe.encode(smiles) if string_format == 'SAFE' else smiles
            st.code(string_to_display)
            # Copy-to-clipboard functionality
            st_copy_button(string_to_display, key=f'copy_{mol_name}')
            # Display ADMET properties
            st.write("**ADMET Properties:**")
            st.write(preds.loc[mol_name])
else:
    st.write("Click the 'Generate Molecules' button to generate beta-lactam molecules.")

# Function Definitions
def generate_molecule_image(input_string, use_safe_visualization=True):
    try:
        if use_safe_visualization:
            try:
                # Attempt to decode as SAFE string
                smiles = safe.decode(input_string)
                # Encode back to SAFE string
                safe_string = safe.encode(smiles)
            except Exception:
                # If decoding fails, assume input is SMILES and encode to SAFE
                safe_string = safe.encode(input_string)
            # Generate SVG image with fragment highlights
            svg_str = safe.to_image(safe_string)
            # Convert SVG to PNG bytes
            png_bytes = cairosvg.svg2png(bytestring=svg_str.encode('utf-8'))
            # Create an image object
            img = Image.open(io.BytesIO(png_bytes))
        else:
            # Generate standard molecule image
            mol = Chem.MolFromSmiles(input_string)
            if mol:
                img = Draw.MolToImage(mol, size=(200, 200))  # Adjusted size
            else:
                img = None
        return img
    except Exception as e:
        # Collect exceptions for later reporting
        return e

import streamlit.components.v1 as components

def st_copy_button(text, key):
    """Creates a copy-to-clipboard button."""
    components.html(f"""
        <button onclick="navigator.clipboard.writeText('{text}')" style="padding:5px;">Copy</button>
    """, height=45)