如何从PDB文件中提取链?

时间:2012-07-27 10:09:49

标签: python bash bioinformatics biopython

我想从pdb文件中提取链。我有一个名为pdb.txt的文件,其中包含pdb ID,如下所示。前四个字符代表PDB ID,最后一个字符是链ID。

1B68A 
1BZ4B
4FUTA

我想1)逐行读取文件                  2)从相应的PDB文件下载每个链的原子坐标                  3)将输出保存到文件夹。

我使用以下脚本来提取链。但是这段代码只打印来自pdb文件的A链。

for i in 1B68 1BZ4 4FUT
do 
wget -c "http://www.pdb.org/pdb/download/downloadFile.do?fileFormat=pdb&compression=NO&structureId="$i -O $i.pdb
grep  ATOM $i.pdb | grep 'A' > $i\_A.pdb
done

3 个答案:

答案 0 :(得分:10)

以下BioPython代码应该很适合您的需求。

它使用PDB.Select仅选择所需的链(在您的情况下,一个链)和PDBIO()来创建仅包含链的结构。

import os
from Bio import PDB


class ChainSplitter:
    def __init__(self, out_dir=None):
        """ Create parsing and writing objects, specify output directory. """
        self.parser = PDB.PDBParser()
        self.writer = PDB.PDBIO()
        if out_dir is None:
            out_dir = os.path.join(os.getcwd(), "chain_PDBs")
        self.out_dir = out_dir

    def make_pdb(self, pdb_path, chain_letters, overwrite=False, struct=None):
        """ Create a new PDB file containing only the specified chains.

        Returns the path to the created file.

        :param pdb_path: full path to the crystal structure
        :param chain_letters: iterable of chain characters (case insensitive)
        :param overwrite: write over the output file if it exists
        """
        chain_letters = [chain.upper() for chain in chain_letters]

        # Input/output files
        (pdb_dir, pdb_fn) = os.path.split(pdb_path)
        pdb_id = pdb_fn[3:7]
        out_name = "pdb%s_%s.ent" % (pdb_id, "".join(chain_letters))
        out_path = os.path.join(self.out_dir, out_name)
        print "OUT PATH:",out_path
        plural = "s" if (len(chain_letters) > 1) else ""  # for printing

        # Skip PDB generation if the file already exists
        if (not overwrite) and (os.path.isfile(out_path)):
            print("Chain%s %s of '%s' already extracted to '%s'." %
                    (plural, ", ".join(chain_letters), pdb_id, out_name))
            return out_path

        print("Extracting chain%s %s from %s..." % (plural,
                ", ".join(chain_letters), pdb_fn))

        # Get structure, write new file with only given chains
        if struct is None:
            struct = self.parser.get_structure(pdb_id, pdb_path)
        self.writer.set_structure(struct)
        self.writer.save(out_path, select=SelectChains(chain_letters))

        return out_path


class SelectChains(PDB.Select):
    """ Only accept the specified chains when saving. """
    def __init__(self, chain_letters):
        self.chain_letters = chain_letters

    def accept_chain(self, chain):
        return (chain.get_id() in self.chain_letters)


if __name__ == "__main__":
    """ Parses PDB id's desired chains, and creates new PDB structures. """
    import sys
    if not len(sys.argv) == 2:
        print "Usage: $ python %s 'pdb.txt'" % __file__
        sys.exit()

    pdb_textfn = sys.argv[1]

    pdbList = PDB.PDBList()
    splitter = ChainSplitter("/home/steve/chain_pdbs")  # Change me.

    with open(pdb_textfn) as pdb_textfile:
        for line in pdb_textfile:
            pdb_id = line[:4].lower()
            chain = line[4]
            pdb_fn = pdbList.retrieve_pdb_file(pdb_id)
            splitter.make_pdb(pdb_fn, chain)

最后一点说明:不要为PDB文件编写自己的解析器。格式规范是丑陋的(非常丑陋),并且错误的PDB文件的数量是惊人的。使用像BioPython这样的工具来处理解析!

此外,您应该使用与PDB数据库交互的工具,而不是使用wget。它们考虑了FTP连接限制,PDB数据库的性质变化等等。我应该知道 - 我updated Bio.PDBList来说明数据库中的变化。 =)

答案 1 :(得分:0)

假设您有以下文件pdb_structures

1B68A 
1BZ4B
4FUTA

然后将您的代码放在load_pdb.sh

while read name
do
    chain=${name:4:1}
    name=${name:0:4}
    wget -c "http://www.pdb.org/pdb/download/downloadFile.do?fileFormat=pdb&compression=NO&structureId="$name -O $name.pdb
    awk -v chain=$chain '$0~/^ATOM/ && substr($0,20,1)==chain {print}' $name.pdb > $name\_$chain.pdb
    #   rm $name.pdb
done

如果您不需要原始pdb,请取消注释最后一行 执行

cat pdb_structures | ./load_pdb.sh

答案 2 :(得分:0)

回答这个问题可能有点晚了,但我会发表意见。 Biopython有一些非常方便的功能可以帮助您轻松实现这样的思考。您可以使用类似自定义选择类的东西,然后为要在原始pdb文件的for循环中选择的每个链调用它。

    from Bio.PDB import Select, PDBIO
    from Bio.PDB.PDBParser import PDBParser

    class ChainSelect(Select):
        def __init__(self, chain):
            self.chain = chain

        def accept_chain(self, chain):
            if chain.get_id() == self.chain:
                return 1
            else:          
                return 0

    chains = ['A','B','C']
    p = PDBParser(PERMISSIVE=1)       
    structure = p.get_structure(pdb_file, pdb_file)

    for chain in chains:
        pdb_chain_file = 'pdb_file_chain_{}.pdb'.format(chain)                                 
        io_w_no_h = PDBIO()               
        io_w_no_h.set_structure(structure)
        io_w_no_h.save('{}'.format(pdb_chain_file), ChainSelect(chain))