多进程蛋白质结构PDB计算接触图

```python
import os
import numpy as np
from Bio.PDB import PDBParser
from multiprocessing import Pool

def calculate_residue_centroid(residue):
    """计算氨基酸残基的几何中心（排除氢原子）"""
    atoms = [atom for atom in residue.get_atoms() if atom.element != 'H']
    coordinates = np.array([atom.get_coord() for atom in atoms])
    return coordinates.mean(axis=0)

def calculate_centroid_contact_map(structure, chain_id, threshold=8.0):
    """生成基于质心的接触矩阵"""
    chain = structure[0][chain_id]
    residues = list(chain.get_residues())
    n = len(residues)
    cmap = np.zeros((n, n), dtype=int)
    
    centroids = [calculate_residue_centroid(res) for res in residues]
    for i in range(n):
        for j in range(i+1, n):
            distance = np.linalg.norm(centroids[i] - centroids[j])
            if distance < threshold:
                cmap[i][j] = cmap[j][i] = 1
    return cmap

def process_file(args):
    """多进程处理单个文件的函数"""
    filename, input_dir, save_dir, threshold = args
    parser = PDBParser()  # 每个进程独立创建解析器
    try:
        pro_name = filename.replace(".pdb", "")
        input_path = os.path.join(input_dir, filename)
        save_path = os.path.join(save_dir, pro_name)
        
        # 解析结构并计算接触矩阵
        structure = parser.get_structure("protein", input_path)
        cmap = calculate_centroid_contact_map(structure, "A", threshold)
        np.save(save_path, cmap)
    except Exception as e:
        print(f"处理文件 {filename} 时出错: {e}")

if __name__ == "__main__":
    # 配置路径和参数
    input_dir = "alphaflod2_pdb/PDB14189/PDB14189"
    save_dir = "alphafold2_map/PDB14189"
    threshold = 10.0
    
    # 确保输出目录存在
    os.makedirs(save_dir, exist_ok=True)
    
    # 获取待处理的PDB文件列表
    filenames = [f for f in os.listdir(input_dir) if f.endswith(".pdb")]
    
    # 构造参数列表（每个元素包含文件信息和参数）
    args_list = [(fname, input_dir, save_dir, threshold) for fname in filenames]
    
    # 创建进程池并行处理
    with Pool() as pool:  # 默认使用所有CPU核心
        pool.map(process_file, args_list)
```