app.py

import streamlit as st
from stmol import showmol
import py3Dmol
import requests
import biotite.structure.io as bsio
import random
import hashlib
import urllib3
from Bio.Blast import NCBIWWW, NCBIXML
from Bio.Seq import Seq
from Bio.SeqRecord import SeqRecord
import time
import urllib.parse

urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning)

st.set_page_config(layout='wide')
st.sidebar.title('🔮 GenPro2 Protein Generator, Structure Predictor, and Analysis Tool')
st.sidebar.write('GenPro2 is an end-to-end protein sequence generator, structure predictor, and analysis tool using [*ESMFold*](https://esmatlas.com/about) and the ESM-2 language model.')

def generate_sequence_from_words(words, length):
    seed = ' '.join(words).encode('utf-8')
    random.seed(hashlib.md5(seed).hexdigest())
    amino_acids = "ACDEFGHIKLMNPQRSTVWY"
    return ''.join(random.choice(amino_acids) for _ in range(length))

def render_mol(pdb):
    pdbview = py3Dmol.view(width=800, height=500)
    pdbview.addModel(pdb, 'pdb')
    pdbview.setStyle({'cartoon': {'color': 'spectrum'}})
    pdbview.setBackgroundColor('white')
    pdbview.zoomTo()
    pdbview.zoom(2, 800)
    pdbview.spin(True)
    showmol(pdbview, height=500, width=800)

def perform_blast_analysis(sequence):
    st.subheader('Protein Analysis')
    with st.spinner("Analyzing generated protein... This may take several minutes. Stay tuned!"):
        progress_bar = st.progress(0)
        
        # Slow down the progress bar
        for i in range(100):
            progress_bar.progress(i + 1)
            time.sleep(0.3)  # Increased sleep time to 0.3 seconds
        
        max_retries = 3
        for attempt in range(max_retries):
            try:
                Entrez.email = "[email protected]"  # Replace with your email
                record = SeqRecord(Seq(sequence), id='random_protein')
                result_handle = NCBIWWW.qblast("blastp", "swissprot", record.seq, expect=10, hitlist_size=1)
                
                blast_record = NCBIXML.read(result_handle)
                
                if blast_record.alignments:
                    alignment = blast_record.alignments[0]  # Get the top hit
                    hsp = alignment.hsps[0]  # Get the first (best) HSP
                    
                    # Extract protein name and organism
                    title_parts = alignment.title.split('|')
                    protein_name = title_parts[-1].strip()
                    organism = title_parts[-2].split('OS=')[-1].split('OX=')[0].strip()
                    
                    # Calculate identity percentage
                    identity_percentage = (hsp.identities / alignment.length) * 100
                    
                    st.write(f"**Top Match:** {protein_name}")
                    st.write(f"**Organism:** {organism}")
                    st.write(f"**Sequence Identity:** {identity_percentage:.2f}%")
                    
                    # Fetch protein function (if available)
                    if hasattr(alignment, 'description') and alignment.description:
                        st.write(f"**Potential Function:** {alignment.description}")
                else:
                    st.write("No significant matches found. This might be a unique protein sequence!")
                
                break  # If successful, exit the retry loop
            
            except Exception as e:
                if attempt < max_retries - 1:
                    st.warning(f"Attempt {attempt + 1} failed. Retrying...")
                    time.sleep(random.uniform(1, 3))  # Add a random delay between retries
                else:
                    st.error(f"An error occurred during protein analysis after {max_retries} attempts: {str(e)}")
                    st.write("Please try again later BLAST could be experiencing server issues.")

def update(sequence, word1, word2, word3, sequence_length):
    headers = {
        'Content-Type': 'application/x-www-form-urlencoded',
    }
    try:
        response = requests.post('https://api.esmatlas.com/foldSequence/v1/pdb/', 
                                 headers=headers, 
                                 data=sequence, 
                                 verify=False,
                                 timeout=300)
        response.raise_for_status()
        pdb_string = response.content.decode('utf-8')
        
        with open('predicted.pdb', 'w') as f:
            f.write(pdb_string)
        
        struct = bsio.load_structure('predicted.pdb', extra_fields=["b_factor"])
        b_value = round(struct.b_factor.mean(), 2)
        
        st.session_state.structure_info = {
            'pdb_string': pdb_string,
            'b_value': b_value,
            'word1': word1,
            'word2': word2,
            'word3': word3,
            'sequence_length': sequence_length
        }
        
        st.session_state.show_analyze_button = True

    except requests.exceptions.RequestException as e:
        st.error(f"An error occurred while calling the API: {str(e)}")
        st.write("Please try again later or contact support if the issue persists.")

def share_on_twitter(word1, word2, word3, length, plddt):
    tweet_text = f"I just generated a new protein using #GenPro2 by @WandsAI from the seed-words '#{word1}', '#{word2}', and '#{word3}' + sequence length of {length}. My Proteins plDDT Score is: {plddt}%!"
    tweet_url = f"https://twitter.com/intent/tweet?text={urllib.parse.quote(tweet_text)}"
    return tweet_url

# Initialize session state variables
if 'sequence' not in st.session_state:
    st.session_state.sequence = None
if 'show_analyze_button' not in st.session_state:
    st.session_state.show_analyze_button = False
if 'structure_info' not in st.session_state:
    st.session_state.structure_info = None

st.title("Word-Seeded Protein Sequence Generator, Structure Predictor, and Analysis Tool")

st.sidebar.subheader("Generate Sequence from Words")
word1 = st.sidebar.text_input("Word 1")
word2 = st.sidebar.text_input("Word 2")
word3 = st.sidebar.text_input("Word 3")
sequence_length = st.sidebar.number_input("Sequence Length", min_value=50, max_value=400, value=100, step=10)

# Information for users
st.info("""
Protein Length Guide:
- 50-100 amino acids: Small proteins/peptides
- 100-300 amino acids: Average protein domains
- 300-500 amino acids: Larger single-domain proteins

""")

if st.sidebar.button('Generate and Predict'):
    if word1 and word2 and word3:
        sequence = generate_sequence_from_words([word1, word2, word3], sequence_length)
        st.session_state.sequence = sequence
        st.sidebar.text_area("Generated Sequence", sequence, height=100)
        st.sidebar.info("Note: The same words and sequence length will always produce the same sequence.")
        
        with st.spinner("Predicting protein structure... This may take a few minutes."):
            update(sequence, word1, word2, word3, sequence_length)
    else:
        st.sidebar.warning("Please enter all three words to generate a sequence.")

# Display structure information if available
if st.session_state.structure_info:
    info = st.session_state.structure_info
    st.subheader(f'Predicted protein structure using seed: {info["word1"]}, {info["word2"]}, and {info["word3"]} + sequence length {info["sequence_length"]}')
    render_mol(info['pdb_string'])
    
    st.subheader('plDDT Confidence Score')
    st.write('plDDT is a bench mark for scoring the confidence level of protein folding predictions based on a scale from 0-100%. 70% or more is good!')
    plddt_score = int(info["b_value"] * 100)
    st.info(f'Your plDDT score is: {plddt_score}%')
 
    st.subheader("Share your unique protein on X(Twitter)")
    
    st.markdown("""
    <div style='background-color: #e6f2ff; padding: 10px; border-radius: 5px; font-size: 0.8em;'>
    <ol>
        <li>Take a screenshot of the protein structure above.</li>
        <li>Click the 'Share Results' link below to open a pre-filled post with your proteins seed-words and plDDT score.</li>
        <li>Be sure to attach a screenshot of your protein before you post!</li>
    </ol>
    </div>
    """, unsafe_allow_html=True)
    
    tweet_url = share_on_twitter(info["word1"], info["word2"], info["word3"], info["sequence_length"], plddt_score)
    st.markdown(f"[Share Results]({tweet_url})")

    st.markdown("""
    ## What to do next:

    """)

    col1, col2 = st.columns(2)
    with col1:
        if st.button('Analyze Protein'):
            perform_blast_analysis(st.session_state.sequence)
    
    with col2:
        st.download_button(
            label="Download PDB",
            data=info['pdb_string'],
            file_name='predicted.pdb',
            mime='text/plain',
        )
        
    st.markdown("""
    If you discover an interesting protein structure, you can explore it even further:
   
    1. Click the 'analyze protein' button to search the [BLAST](https://blast.ncbi.nlm.nih.gov/Blast.cgi?PROGRAM=blastp&PAGE_TYPE=BlastSearch&LINK_LOC=blasthome) protein database and see if your protein matches any known sequences.  The sequence identity will show how close your sequence matches. *Note this can take several minutes
    2. Download your protein data and visit the [Protein Data Bank (PDB)](https://www.rcsb.org/) to match your protein structure against known protein structures.
    3. If you think you've discovered a new and useful protein for humans message us!

  
    **Remember, this folding is based on randomly generated sequences. Interpret the results with caution.
    Enjoy exploring the world of protein sequences!
    """)