python之将具有多个结构的PDB文件解析为数组

我有一个包含几千个结构的 PDB 文件，我想将前十个结构的 alpha 碳的位置坐标保存到一个 numpy 数组中。我可以使用下面的代码将具有单一结构的 PDB 文件解析为数组，但不能将其扩展到具有多个结构的文件。

from Bio.PDB.PDBParser import PDBParser 
import numpy 
 
pdb_filename ='./1fqy.pdb' 
parser = PDBParser(PERMISSIVE=1) 
structure = parser.get_structure("1fqy", pdb_filename) 
model = structure[0] 
chain = model["A"] 
 
S1coor = numpy.zeros(shape=(226, 3)) 
i = 0 
 
for residue1 in chain: 
     resnum = residue1.get_id()[1] 
     atom1 = residue1['CA'] 
     S1coor[i] = atom1.get_coord() 
     i = i + 1 

请您参考如下方法:

from Bio.PDB.PDBParser import PDBParser 
import numpy , tempfile ,os , re 
 
models_re = re.compile("MODEL") 
pdb_re = re.compile(r"MODEL(.*?)ENDMDL", re.DOTALL) 
 
def PDB_parse(pdb_file_handle): 
    model_pos = [] 
    models = [] 
    k = open(pdb_file_handle,"r").read() 
    for i in models_re.finditer(k): 
        model_pos.append(i.start()) 
    for i in model_pos: 
        models.append(pdb_re.search(k,i).group()) 
    return models 
 
array_all_structure = [] 
 
for i in PDB_parse(pdb_file_handle): 
    temp_file = tempfile.NamedTemproaryFile(delete = False) 
    temp_file.write(i) 
    temp_file.close 
    structure = parser.get_structure("1fqy", temp_file.name) 
    os.remove(temp_file.name) 
    model = structure[0] 
    chain = model["A"] 
    S1coor = numpy.zeros(shape=(226, 3)) 
    i = 0 
    for residue1 in chain: 
       resnum = residue1.get_id()[1] 
       atom1 = residue1['CA'] 
       S1coor[i] = atom1.get_coord() 
       i = i + 1 
       array_all_structure.append(i) 

可能这种链接器会有所帮助，您首先隔离 pdb 文件，然后相应地读取它们。