app.py

import torch
import random
import pandas as pd
from utils import create_vocab, setup_seed
from dataset_mlm import  get_paded_token_idx_gen, add_tokens_to_vocab
import gradio as gr
from gradio_rangeslider import RangeSlider
import time

is_stopped = False

def temperature_sampling(logits, temperature):
    logits = logits / temperature
    probabilities = torch.softmax(logits, dim=-1)
    sampled_token = torch.multinomial(probabilities, 1)
    return sampled_token

def stop_generation():
    global is_stopped
    is_stopped = True
    return "Generation stopped."

def CTXGen(X1, X2, τ, g_num, length_range, model_name, seed):
    if seed =='random':
        seed = random.randint(0,100000)
        setup_seed(seed)
    else:
        setup_seed(int(seed))
        
    global is_stopped
    is_stopped = False
    start, end = length_range

    device = torch.device("cpu")
    vocab_mlm = create_vocab()
    vocab_mlm = add_tokens_to_vocab(vocab_mlm)
    save_path = model_name
    train_seqs = pd.read_csv('C0_seq.csv')
    train_seq = train_seqs['Seq'].tolist()
    model = torch.load(save_path, map_location=torch.device('cpu'))
    model = model.to(device)

    msa_data = pd.read_csv('conoData_C0.csv') 
    msa = msa_data['Sequences'].tolist()
    msa = [x for x in msa if x.startswith(f"{X1}|{X2}")]
    msa = random.choice(msa)
    X4 = msa.split("|")[3]
    X5 = msa.split("|")[4]
    X6 = msa.split("|")[5]

    model.eval()
    with torch.no_grad():
        IDs = []
        generated_seqs = []
        generated_seqs_FINAL = []
        cls_probability_all = []
        act_probability_all = []

        count = 0
        gen_num = int(g_num)
        NON_AA = ["B","O","U","Z","X",'<K16>', '<α1β1γδ>', '<Ca22>', '<AChBP>', '<K13>', '<α1BAR>', '<α1β1ε>', '<α1AAR>', '<GluN3A>', '<α4β2>',
                        '<GluN2B>', '<α75HT3>', '<Na14>', '<α7>', '<GluN2C>', '<NET>', '<NavBh>', '<α6β3β4>', '<Na11>', '<Ca13>', 
                        '<Ca12>', '<Na16>', '<α6α3β2>', '<GluN2A>', '<GluN2D>', '<K17>', '<α1β1δε>', '<GABA>', '<α9>', '<K12>', 
                        '<Kshaker>', '<α3β4>', '<Na18>', '<α3β2>', '<α6α3β2β3>', '<α1β1δ>', '<α6α3β4β3>', '<α2β2>','<α6β4>', '<α2β4>',
                        '<Na13>', '<Na12>', '<Na15>', '<α4β4>', '<α7α6β2>', '<α1β1γ>', '<NaTTXR>', '<K11>', '<Ca23>', 
                        '<α9α10>','<α6α3β4>', '<NaTTXS>', '<Na17>','<high>','<low>','[UNK]','[SEP]','[PAD]','[CLS]','[MASK]']             
        start_time = time.time()
        while count < gen_num:
            new_seq = None
            if is_stopped:
                return "output.csv", pd.DataFrame()

            if time.time() - start_time > 1200:
                break

            gen_len = random.randint(int(start), int(end))
            X3 = "X" * gen_len
            seq = [f"{X1}|{X2}|{X3}|{X4}|{X5}|{X6}"]
            vocab_mlm.token_to_idx["X"] = 4

            padded_seq, _, _, _ = get_paded_token_idx_gen(vocab_mlm, seq, new_seq)
            input_text = ["[MASK]" if i=="X" else i for i in padded_seq]

            gen_length = len(input_text)
            length = gen_length - sum(1 for x in input_text if x != '[MASK]')

            for i in range(length):
                if is_stopped:
                    return "output.csv", pd.DataFrame()
                
                _, idx_seq, idx_msa, attn_idx = get_paded_token_idx_gen(vocab_mlm, seq, new_seq)
                idx_seq = torch.tensor(idx_seq).unsqueeze(0).to(device)
                idx_msa = torch.tensor(idx_msa).unsqueeze(0).to(device)
                attn_idx = torch.tensor(attn_idx).to(device)

                mask_positions = [j for j in range(gen_length) if input_text[j] == "[MASK]"]
                mask_position = torch.tensor([mask_positions[torch.randint(len(mask_positions), (1,))]])
                
                logits = model(idx_seq,idx_msa, attn_idx) 
                mask_logits = logits[0, mask_position.item(), :] # 

                predicted_token_id = temperature_sampling(mask_logits, τ)

                predicted_token = vocab_mlm.to_tokens(int(predicted_token_id))
                input_text[mask_position.item()] = predicted_token
                padded_seq[mask_position.item()] = predicted_token.strip()
                new_seq = padded_seq

            generated_seq = input_text
        
            generated_seq[1] = "[MASK]"
            input_ids = vocab_mlm.__getitem__(generated_seq)
            logits = model(torch.tensor([input_ids]).to(device), idx_msa)
            cls_mask_logits = logits[0, 1, :]
            cls_probability, cls_mask_probs = torch.topk((torch.softmax(cls_mask_logits, dim=-1)), k=85)

            generated_seq[2] = "[MASK]"
            input_ids = vocab_mlm.__getitem__(generated_seq)
            logits = model(torch.tensor([input_ids]).to(device), idx_msa)
            act_mask_logits = logits[0, 2, :]
            act_probability, act_mask_probs = torch.topk((torch.softmax(act_mask_logits, dim=-1)), k=2)

            cls_pos = vocab_mlm.to_tokens(list(cls_mask_probs))
            act_pos = vocab_mlm.to_tokens(list(act_mask_probs))
            
            if X1 in cls_pos and X2 in act_pos:
                cls_proba = cls_probability[cls_pos.index(X1)].item()
                act_proba = act_probability[act_pos.index(X2)].item()
                generated_seq = generated_seq[generated_seq.index('[MASK]') + 2:generated_seq.index('[SEP]')]
                if act_proba>=0.5 and generated_seq.count('C') % 2 == 0 and len("".join(generated_seq)) == gen_len:
                    generated_seqs.append("".join(generated_seq))
                    if "".join(generated_seq) not in train_seq and "".join(generated_seq) not in generated_seqs[0:-1] and all(x not in NON_AA for x in generated_seq):
                        generated_seqs_FINAL.append("".join(generated_seq))
                        cls_probability_all.append(cls_proba)
                        act_probability_all.append(act_proba)
                        IDs.append(count+1)
                        out = pd.DataFrame({
                            'ID':IDs,
                            'Generated_seq': generated_seqs_FINAL,
                            'Subtype': X1,
                            'Subtype_probability': cls_probability_all, 
                            'Potency': X2, 
                            'Potency_probability': act_probability_all, 
                            'Random_seed': seed
                        })
                        out.to_csv("output.csv", index=False, encoding='utf-8-sig')
                        count += 1
                        yield "output.csv", out
    return "output.csv", out

with gr.Blocks() as demo:
    gr.Markdown("🔗 **[Label Prediction](https://huggingface.co/spaces/oucgc1996/CreoPep_Label_Prediction)**"
                "🔗 **[Unconstrained Generation](https://huggingface.co/spaces/oucgc1996/CreoPep_Unconstrained_generation)**"
                "🔗 **[Conditional Generation](https://huggingface.co/spaces/oucgc1996/CreoPep_conditional_generation)**"
                "🔗 **[Optimization Generation](https://huggingface.co/spaces/oucgc1996/CreoPep_optimization_generation)**")
    gr.Markdown("# Conotoxin Conditional Generation")
    gr.Markdown("#### Input")
    gr.Markdown("✅**Subtype**: subtype of action. For example, α7.")
    gr.Markdown("✅**Potency**: required potency. For example, High.")
    gr.Markdown("✅**τ**: temperature factor controls the diversity of conotoxins generated. The higher the value, the higher the diversity.")
    gr.Markdown("✅**Number of generations**: if it is not completed within 1200 seconds, it will automatically stop.")
    gr.Markdown("✅**Length range**: expected length range of conotoxins generated.")
    gr.Markdown("✅**Model**: model parameters trained at different stages of data augmentation. Please refer to the paper for details.")
    gr.Markdown("✅**Seed**: enter an integer as the random seed to ensure reproducible results. The default is random.")
    
    with gr.Row():
        X1 = gr.Dropdown(choices=['<α7>', '<AChBP>', '<Ca12>', '<Ca13>', '<Ca22>', '<Ca23>', '<GABA>', '<GluN2A>', '<GluN2B>', '<GluN2C>', '<GluN2D>', '<GluN3A>', 
                             '<K11>', '<K12>', '<K13>', '<K16>', '<K17>', '<Kshaker>',
                             '<Na11>', '<Na12>', '<Na13>', '<Na14>', '<Na15>', '<Na16>', '<Na17>', '<Na18>', '<NaTTXR>', '<NaTTXS>', '<NavBh>', '<NET>', 
                             '<α1AAR>', '<α1BAR>', '<α1β1γ>', '<α1β1γδ>', '<α1β1δ>', '<α1β1δε>', '<α1β1ε>', '<α2β2>', '<α2β4>', '<α3β2>', '<α3β4>', 
                             '<α4β2>', '<α4β4>', '<α6α3β2>', '<α6α3β2β3>', '<α6α3β4>', '<α6α3β4β3>', '<α6β3β4>', '<α6β4>', '<α7α6β2>', 
                             '<α75HT3>', '<α9>', '<α9α10>'], label="Subtype")
        X2 = gr.Dropdown(choices=['<high>','<low>'], label="Potency")
        τ = gr.Slider(minimum=1, maximum=2, step=0.1, label="τ")
        g_num = gr.Dropdown(choices=[1, 10, 20, 30, 40, 50], label="Number of generations")
    with gr.Row():
        length_range = RangeSlider(minimum=8, maximum=50, step=1, value=(12, 16), label="Length range")
        model_name = gr.Dropdown(choices=['model_final.pt','model_C1.pt','model_C2.pt','model_C3.pt','model_C4.pt','model_C5.pt','model_mlm.pt'], label="Model")
        seed = gr.Textbox(label="Seed", value="random")
    with gr.Row():
        start_button = gr.Button("Start Generation")
        stop_button = gr.Button("Stop Generation")
    with gr.Row():
        output_file = gr.File(label="Download generated conotoxins")
    with gr.Row():
        output_df = gr.DataFrame(label="Generated Conotoxins")
    
    start_button.click(CTXGen, inputs=[X1, X2, τ, g_num, length_range,model_name,seed], outputs=[output_file, output_df])
    stop_button.click(stop_generation, outputs=None)

demo.launch()