sample_diffusion.py

import argparse
import os
import shutil
import time
import sys
sys.path.append(os.path.abspath('./'))

import numpy as np
import torch
from torch_geometric.data import Batch
from torch_geometric.transforms import Compose
from torch_scatter import scatter_sum, scatter_mean
from tqdm.auto import tqdm

import utils.misc as misc
import utils.transforms as trans
from datasets import get_dataset
from datasets.pl_data import FOLLOW_BATCH
from models.molopt_score_model import ScorePosNet3D, log_sample_categorical
from utils.evaluation import atom_num


def unbatch_v_traj(ligand_v_traj, n_data, ligand_cum_atoms):
    all_step_v = [[] for _ in range(n_data)]
    for v in ligand_v_traj:  # step_i
        v_array = v.cpu().numpy()
        for k in range(n_data):
            all_step_v[k].append(v_array[ligand_cum_atoms[k]:ligand_cum_atoms[k + 1]])
    all_step_v = [np.stack(step_v) for step_v in all_step_v]  # num_samples * [num_steps, num_atoms_i]
    return all_step_v


def sample_diffusion_ligand(model, data, num_samples, batch_size=16, device='cuda',
                            num_steps=None, center_pos_mode='protein',
                            sample_num_atoms='prior',guide_mode='joint',
                            value_model=None,
        type_grad_weight=1.,pos_grad_weight=1.):
    
    all_pred_pos, all_pred_v, all_pred_exp = [], [], []
    all_pred_pos_traj, all_pred_v_traj, all_pred_exp_traj, all_pred_exp_atom_traj = [], [], [], []
    all_pred_v0_traj, all_pred_vt_traj = [], []
    time_list = []
    num_batch = int(np.ceil(num_samples / batch_size))
    current_i = 0
    for i in tqdm(range(num_batch)):
        n_data = batch_size if i < num_batch - 1 else num_samples - batch_size * (num_batch - 1)
        batch = Batch.from_data_list([data.clone() for _ in range(n_data)], follow_batch=FOLLOW_BATCH).to(device)

        t1 = time.time()
        with torch.no_grad():
            batch_protein = batch.protein_element_batch
            if sample_num_atoms == 'prior':
                pocket_size = atom_num.get_space_size(batch.protein_pos.detach().cpu().numpy())
                ligand_num_atoms = [atom_num.sample_atom_num(pocket_size).astype(int) for _ in range(n_data)]
                batch_ligand = torch.repeat_interleave(torch.arange(n_data), torch.tensor(ligand_num_atoms)).to(device)
            elif sample_num_atoms == 'range':
                ligand_num_atoms = list(range(current_i + 1, current_i + n_data + 1))
                batch_ligand = torch.repeat_interleave(torch.arange(n_data), torch.tensor(ligand_num_atoms)).to(device)
            elif sample_num_atoms == 'ref':
                batch_ligand = batch.ligand_element_batch
                ligand_num_atoms = scatter_sum(torch.ones_like(batch_ligand), batch_ligand, dim=0).tolist()
            else:
                raise ValueError

            # init ligand pos
            center_pos = scatter_mean(batch.protein_pos, batch_protein, dim=0)
            batch_center_pos = center_pos[batch_ligand]
            init_ligand_pos = batch_center_pos + torch.randn_like(batch_center_pos)

            # init ligand v
            uniform_logits = torch.zeros(len(batch_ligand), model.num_classes).to(device)
            init_ligand_v_prob = log_sample_categorical(uniform_logits)
            init_ligand_v = init_ligand_v_prob.argmax(dim=-1)

            r = model.sample_diffusion(
                guide_mode=guide_mode,
                value_model=value_model,
                type_grad_weight=type_grad_weight,
                pos_grad_weight=pos_grad_weight,
                protein_pos=batch.protein_pos,
                protein_v=batch.protein_atom_feature.float(),
                batch_protein=batch_protein,

                init_ligand_pos=init_ligand_pos,
                init_ligand_v=init_ligand_v,
                batch_ligand=batch_ligand,
                num_steps=num_steps,
                center_pos_mode=center_pos_mode
            )
            ligand_pos, ligand_v, ligand_pos_traj, ligand_v_traj = r['pos'], r['v'], r['pos_traj'], r['v_traj']
            ligand_v0_traj, ligand_vt_traj = r['v0_traj'], r['vt_traj']
            exp_traj = r['exp_traj']
            exp_atom_traj = r['exp_atom_traj']
            # unbatch exp
            if guide_mode == 'joint' or guide_mode == 'pdbbind_random' or guide_mode == 'valuenet' or guide_mode == 'wo':
                all_pred_exp += exp_traj[-1]
                all_pred_exp_traj += exp_traj
            
            # unbatch pos
            ligand_cum_atoms = np.cumsum([0] + ligand_num_atoms)
            ligand_pos_array = ligand_pos.cpu().numpy().astype(np.float64)
            all_pred_pos += [ligand_pos_array[ligand_cum_atoms[k]:ligand_cum_atoms[k + 1]] for k in
                             range(n_data)]  # num_samples * [num_atoms_i, 3]

            all_step_pos = [[] for _ in range(n_data)]
            for p in ligand_pos_traj:  # step_i
                p_array = p.cpu().numpy().astype(np.float64)
                for k in range(n_data):
                    all_step_pos[k].append(p_array[ligand_cum_atoms[k]:ligand_cum_atoms[k + 1]])
            all_step_pos = [np.stack(step_pos) for step_pos in
                            all_step_pos]  # num_samples * [num_steps, num_atoms_i, 3]
            all_pred_pos_traj += [p for p in all_step_pos]

            # unbatch v
            ligand_v_array = ligand_v.cpu().numpy()
            all_pred_v += [ligand_v_array[ligand_cum_atoms[k]:ligand_cum_atoms[k + 1]] for k in range(n_data)]

            all_step_v = unbatch_v_traj(ligand_v_traj, n_data, ligand_cum_atoms)
            all_pred_v_traj += [v for v in all_step_v]
            all_step_v0 = unbatch_v_traj(ligand_v0_traj, n_data, ligand_cum_atoms)
            all_pred_v0_traj += [v for v in all_step_v0]
            all_step_vt = unbatch_v_traj(ligand_vt_traj, n_data, ligand_cum_atoms)
            all_pred_vt_traj += [v for v in all_step_vt]
            all_step_exp_atom = unbatch_v_traj(exp_atom_traj, n_data, ligand_cum_atoms)
            all_pred_exp_atom_traj += [v for v in all_step_exp_atom]
            
            
        t2 = time.time()
        time_list.append(t2 - t1)
        current_i += n_data
        
        
    all_pred_exp = torch.stack(all_pred_exp,dim=0).numpy()
    all_pred_exp_traj = torch.stack(all_pred_exp_traj,dim=0).numpy()
        
    return all_pred_pos, all_pred_v, all_pred_exp, all_pred_pos_traj, all_pred_v_traj, all_pred_exp_traj, all_pred_v0_traj, all_pred_vt_traj, all_pred_exp_atom_traj, time_list

def main():
    parser = argparse.ArgumentParser()
    parser.add_argument('--config', type=str, default='./configs/sampling.yml')
    parser.add_argument('-i', '--data_id', type=int, default=81)
    parser.add_argument('--device', type=str, default='cuda')
    parser.add_argument('--batch_size', type=int, default=4)
    parser.add_argument('--guide_mode', type=str, default='joint', choices=['joint', 'pdbbind_random', 'vina', 'valuenet', 'wo'])  
    parser.add_argument('--type_grad_weight', type=float, default=0)
    parser.add_argument('--pos_grad_weight', type=float, default=0)
    parser.add_argument('--result_path', type=str, default='./test_package')
    if len(sys.argv[1:]) == 0:
        parser.print_help()
        exit()
    args = parser.parse_args()
        
    result_path = args.result_path
    os.makedirs(result_path, exist_ok=True)
    shutil.copyfile(args.config, os.path.join(result_path, 'sample.yml'))
    logger = misc.get_logger('sampling', log_dir=result_path)

    # Load config
    config = misc.load_config(args.config)
    logger.info(config)
    misc.seed_all(config.sample.seed)

    # Load checkpoint
    if args.guide_mode == 'joint':
        ckpt = torch.load(config.model['joint_ckpt'], map_location=args.device)
        value_ckpt = None
    elif args.guide_mode == 'pdbbind_random':
        ckpt = torch.load(config.model['pdbbind_random'], map_location=args.device)
        value_ckpt = None
    elif args.guide_mode == 'vina':
        ckpt = torch.load(config.model['policy_ckpt'], map_location=args.device)
        value_ckpt = None
    elif args.guide_mode == 'valuenet':
        ckpt = torch.load(config.model['policy_ckpt'], map_location=args.device)
        value_ckpt = torch.load(config.model['value_ckpt'], map_location=args.device)
    elif args.guide_mode == 'wo':
        ckpt = torch.load(config.model['policy_ckpt'], map_location=args.device)
        value_ckpt = None
    else:
        raise NotImplementedError
    
    logger.info(f"Training Config: {ckpt['config']}")
    logger.info(f"args: {args}")
    
    # Transforms
    protein_featurizer = trans.FeaturizeProteinAtom()
    ligand_atom_mode = ckpt['config'].data.transform.ligand_atom_mode
    ligand_featurizer = trans.FeaturizeLigandAtom(ligand_atom_mode)
    transform = Compose([
        protein_featurizer,
        ligand_featurizer,
        trans.FeaturizeLigandBond(),
    ])

    # Load dataset
    dataset, subsets = get_dataset(
        config=ckpt['config'].data,
        transform=transform
    )
    if ckpt['config'].data.name == 'pl':
        test_set = subsets['test']
    elif ckpt['config'].data.name == 'pdbbind':
        test_set = subsets['test']
    else:
        raise ValueError
    logger.info(f'Test: {len(test_set)}')


    # Load model
    model = ScorePosNet3D(
        ckpt['config'].model,
        protein_atom_feature_dim=protein_featurizer.feature_dim,
        ligand_atom_feature_dim=ligand_featurizer.feature_dim
    ).to(args.device)
    model.load_state_dict(ckpt['model'])
    
    if value_ckpt is not None:
        # value model
        value_model = ScorePosNet3D(
            value_ckpt['config'].model,
            protein_atom_feature_dim=protein_featurizer.feature_dim,
            ligand_atom_feature_dim=ligand_featurizer.feature_dim
        ).to(args.device)
        value_model.load_state_dict(value_ckpt['model'])
    else:
        value_model = None
        

    data = test_set[args.data_id]
    pred_pos, pred_v, pred_exp, pred_pos_traj, pred_v_traj, pred_exp_traj, pred_v0_traj, pred_vt_traj, pred_exp_atom_traj, time_list = sample_diffusion_ligand(
        model, data, config.sample.num_samples,
        batch_size=args.batch_size, device=args.device,
        num_steps=config.sample.num_steps,
        center_pos_mode=config.sample.center_pos_mode,
        sample_num_atoms=config.sample.sample_num_atoms,
        guide_mode=args.guide_mode,
        value_model=value_model,
        type_grad_weight=args.type_grad_weight,
        pos_grad_weight=args.pos_grad_weight
    )
    result = {
        'data': data,
        'pred_ligand_pos': pred_pos,
        'pred_ligand_v': pred_v,
        'pred_exp': pred_exp,
        'pred_ligand_pos_traj': pred_pos_traj,
        'pred_ligand_v_traj': pred_v_traj,
        'pred_exp_traj': pred_exp_traj,
        'pred_exp_atom_traj': pred_exp_atom_traj,
        'time': time_list
    }
    logger.info('Sample done!')

    
    torch.save(result, os.path.join(result_path, f'result_{args.data_id}.pt'))


if __name__ == '__main__':
    main()