app.py

"""
ZeroGPU Structure Prediction API
"""

import spaces
import gradio as gr
import torch
from transformers import EsmForProteinFolding, AutoTokenizer

print("Loading ESMFold model...")
MODEL_NAME = "facebook/esmfold_v1"

tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
model = EsmForProteinFolding.from_pretrained(MODEL_NAME)

if torch.cuda.is_available():
    model = model.cuda()
    model.esm = model.esm.half()
    print(f"Model loaded on GPU: {torch.cuda.get_device_name(0)}")
else:
    print("Model loaded on CPU")


@spaces.GPU(duration=120)
def predict_structure(sequence: str) -> str:
    sequence = sequence.strip().upper()
    valid_aa = set("ACDEFGHIKLMNPQRSTVWY")
    
    if not sequence:
        return "Error: Empty sequence provided"
    
    invalid_chars = set(sequence) - valid_aa
    if invalid_chars:
        return f"Error: Invalid amino acids found: {invalid_chars}"
    
    if len(sequence) > 500:
        return "Error: Sequence too long (max 500 residues)"
    
    try:
        inputs = tokenizer(sequence, return_tensors="pt", add_special_tokens=False)
        
        if torch.cuda.is_available():
            inputs = {k: v.cuda() for k, v in inputs.items()}
        
        with torch.no_grad():
            outputs = model(**inputs)
        
        pdb_string = model.output_to_pdb(outputs)[0]
        return pdb_string
        
    except Exception as e:
        return f"Error: {str(e)}"


with gr.Blocks(title="🧬 Antibody Structure API") as demo:
    gr.Markdown("""
    # 🧬 Antibody Structure Prediction API (ZeroGPU)
    
    GPU-accelerated ESMFold structure prediction.
    
    **API Usage:**
    ```python
    from gradio_client import Client
    client = Client("kmlyyll/antibody-structure-api")
    pdb = client.predict(sequence, api_name="/predict")
    ```
    """)
    
    seq_input = gr.Textbox(label="Amino Acid Sequence", placeholder="Enter sequence...", lines=3)
    predict_btn = gr.Button("Predict Structure", variant="primary")
    pdb_output = gr.Textbox(label="PDB Output", lines=20)
    
    predict_btn.click(fn=predict_structure, inputs=seq_input, outputs=pdb_output, api_name="predict")

if __name__ == "__main__":
    demo.launch()