LightDock.server Linux 双跑比较

原创已于 2025-09-19 16:44:52 修改 · 325 阅读

9 ·

CC 4.0 BY-SA版权

文章标签：

#计算化学 #计算生物

于 2025-09-05 18:03:01 首次发布

LightDock: a new multi-scale approach to protein–protein docking

The LightDock server is free and open to all users and there is no login requirement

server

1示例

故去除约束 next step

结果有正有负合理

2.常见警告⚠

Structure contains HETATM entries. Please remove them or rename to HETATM if possible.

蛋白pdb文件删除杂原子 HETATM；保留或去除水分子

Error reading provided PDB structure

计算化学公社 - 高水平计算化学、理论化学交流论坛

结果：

DNA图像未出现

转换dome 可将DS visualizer .pdb 文件转换为 lightdock.server需要的文件

import sys
from pathlib import Path


def normalize_atom_name(raw_name: str) -> str:
    name = raw_name.replace('"', '').strip()
    # Rename phosphate oxygens to example style
    if name == 'OP1':
        name = 'O1P'
    elif name == 'OP2':
        name = 'O2P'
    elif name == 'OP3':
        # Example file does not use OP3; map to O3P if present
        name = 'O3P'
    return name


def format_atom_line(
    serial: int,
    name: str,
    resname: str,
    chain_id: str,
    resseq: int,
    inscode: str,
    x: float,
    y: float,
    z: float,
) -> str:
    # PDB fixed-width formatting approximating the example file layout
    # Columns: Record(1-6), serial(7-11), atom name(13-16), resname(18-20), chain(22), resseq(23-26),
    # coords x(31-38), y(39-46), z(47-54), occupancy(55-60), temp(61-66)
    # Atom name must be right-justified in columns 13-16. Column 17 (altLoc) is a space.
    name_field = name.rjust(4)[:4]
    res_field = resname.rjust(3)[:3]
    chain_field = (chain_id or 'B')[:1]
    ins_field = (inscode or ' ')[:1]
    return (
        # Insert an explicit altLoc space between atom name and residue name
        f"ATOM{serial:7d} {name_field} {res_field} {chain_field}{resseq:4d}{ins_field}"
        f"   {x:8.3f}{y:8.3f}{z:8.3f}  1.00  0.00"
    )


def convert_pdb(in_path: Path, out_path: Path, target_chain: str = 'B') -> None:
    lines_out = []
    serial = 1

    with in_path.open('r', encoding='utf-8', errors='ignore') as f:
        for raw in f:
            record = raw[:6].strip()
            if record not in {"ATOM", "HETATM"}:
                # Skip headers like REMARK, CRYST1, etc. Keep TER and END as we will add later.
                continue

            # Robustly parse using PDB column positions when available
            # Fallback to whitespace splitting if line is too short
            try:
                name = raw[12:16]
                resname = raw[17:20].strip() or 'UNK'
                chain_id = target_chain or (raw[21:22].strip() or 'B')
                resseq_str = raw[22:26].strip() or '0'
                inscode = raw[26:27]
                x = float(raw[30:38])
                y = float(raw[38:46])
                z = float(raw[46:54])
            except Exception:
                parts = raw.split()
                if len(parts) < 8:
                    continue
                # parts: ATOM serial name resname chain resseq x y z ...
                name = parts[2]
                resname = parts[3]
                chain_id = target_chain or (parts[4] if len(parts) > 4 else 'B')
                resseq_str = parts[5] if len(parts) > 5 else '0'
                x, y, z = map(float, parts[6:9])
                inscode = ' '

            # Normalize fields
            name = normalize_atom_name(name)
            # Ensure sugar atom naming keeps apostrophes, e.g., O3', O5'
            # Already handled by stripping quotes only

            try:
                resseq = int(resseq_str)
            except Exception:
                resseq = 0

            # Build output line
            out_line = format_atom_line(
                serial=serial,
                name=name,
                resname=resname,
                chain_id=target_chain or chain_id,
                resseq=resseq,
                inscode=inscode,
                x=x,
                y=y,
                z=z,
            )
            lines_out.append(out_line)
            serial += 1

    # Append TER and END like the example
    if lines_out:
        # Try to infer last residue info for TER line formatting
        last = lines_out[-1]
        # Extract fields to place a TER similar to example (not strictly column-perfect)
        ter_chain = (target_chain or 'B')[:1]
        # Residue name and number from last line columns
        resname = last[17:20]
        resseq = last[22:26].strip()
        lines_out.append(f"TER{''.ljust(8)}      {resname} {ter_chain}{resseq:>4}")
    lines_out.append("END")

    out_path.write_text("\n".join(lines_out) + "\n", encoding='utf-8')


def main():
    if len(sys.argv) < 3:
        print(
            "Usage: python convert_to_example_format.py <input_pdb> <output_pdb> [chain_id]",
        )
        sys.exit(1)

    in_file = Path(sys.argv[1])
    out_file = Path(sys.argv[2])
    chain = sys.argv[3] if len(sys.argv) > 3 else 'B'

    if not in_file.exists():
        print(f"Input file not found: {in_file}")
        sys.exit(1)

    out_file.parent.mkdir(parents=True, exist_ok=True)
    convert_pdb(in_file, out_file, target_chain=chain)
    print(f"Converted {in_file.name} -> {out_file}")


if __name__ == '__main__':
    main()

Run: 6K4J-13-converted

Flexible backbone (双√)

Flexible backbone 2 （单√2）

Flexible backbone 2 （单√1）

Run: 6k4j_10_c_f_1

Linux

上传原始文件

Protonation 质子化

在生物化学和分子生物学中，质子化通常指的是分子或原子团接受一个质子（H+）的过程，这在蛋白质的结构和功能中尤为重要。

Protein

reduce -Trim 1AZP_A.pdb > 1AZP_A_noh.pdb
reduce -BUILD 1AZP_A_noh.pdb > 1AZP_A_h.pdb

DNA

reduce -Trim 1AZP_B.pdb > 1AZP_B_noh.pdb
reduce -BUILD 1AZP_B_noh.pdb > 1AZP_B_h.pdb

1.上传文件reduce_to_amber

python reduce_to_amber.py 1AZP_B_h.pdb dna.pdb

3.Atom not found in mapping: HO3' 14.044 -6.401 -11.426
Atom not found in mapping: HO3' -11.299 10.225 -9.235

Setup

上传文件 restraints.list

lightdock3_setup.py protein.pdb dna.pdb -anm -rst restraints.list

Simulation

上传文件 setup.json

lightdock3.py setup.json 100 -s dna -c 8

上传analyze.sh

./analyze.sh

理查森实验室

LightDock Server

RCSB PDB - 6K4J: Crystal Structure of the the CD9

计算化学公社 - 高水平计算化学、理论化学交流论坛