Building a Protein Data Bank (PDB) file

Question

I have to model a complex polymer of lysine. It resembles a protein, except the lysines don't always bound with their alpha amine. My goal is to produce a PDB (Protein Data Bank) for further calculations.

Do you know any module allowing me to build a molecule? I have special needs, like:

• I have chain of amino acid
• I need to add one more to the chain
• but I have to be able to specify where and how this n+1 amino acid bounds to the previous one
• At the end, I have to be able to generate a PDB file

My first approach was to use the SMILES format, and I can do everything perfectly, except I can not build a pdb at the end, because there is no software capable of handling my number of atoms (> 15000).

Answer 1

Ok, I finally found a solution to my problem, even if it's not related to my first question. I followed my first approach to generate my polymer: a python script generating a long SMILE chain. And then, I finally found a software capable of converting such a long chain: Discovery studio visualizer.

I think it works because this software is not based on openbabel. It's not a Python solution, but it does the work.

Thanks anyway.

Answer 2

Off-hand, I know of two relatively good modules:

pdb-tools

pdbTools is a set of command line python scripts that manipulate wwPDB protein and nucleic acid structure files. There are many programs, both open source and proprietary, that perform similar tasks; however, most of these tools are buried within programs of larger functionality. Thus, relatively simple calculations often involve learning a new program, compiling modules, and installing libraries. To fill a niche (and get the tasks done that I needed done), I started writing my own toolset. This has evolved into the pdbTools suite. The suite of programs is characterized by the following philosophy:

• Each program should run as a stand-alone application with a standard, GNU/POSIX style command line interface.
• Each program should be written in such a way to allow it to be used as a library of functions for more complex programs.
• Programs should require a minimum of external dependencies.

biopython

The Biopython Project is an international association of developers of freely available Python tools for computational molecular biology.

Answer 3

You could overlay selected atoms from the amino acid and the protein to make an artificial bond in pymol - https://pymolwiki.org/index.php/Pair_fit

• then export as a .pdb

you can have a play with the molecule/s positions before pairing.