Spaces:
Sleeping
Sleeping
File size: 8,929 Bytes
fb8c550 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 | import asyncio
import logging
import os
import re
import shutil
import tempfile
import time
from typing import Any
import httpx
from app.tools.base import BaseTool
logger = logging.getLogger(__name__)
PDB_DOWNLOAD = "https://files.rcsb.org/download/{pdb_id}.pdb"
VINA_CMD = shutil.which("vina") or "/usr/local/bin/vina"
def _find_ligand_center(pdb_content: str) -> tuple[float, float, float] | None:
"""Find the geometric center of the largest HETATM ligand (non-water)."""
het_atoms: list[list[tuple[float, float, float]]] = []
current_het: list[tuple[float, float, float]] = []
current_resname = ""
for line in pdb_content.splitlines():
if line.startswith("HETATM"):
resname = line[17:20].strip()
if resname == "HOH":
continue
try:
x = float(line[30:38].strip())
y = float(line[38:46].strip())
z = float(line[46:54].strip())
except ValueError:
continue
if resname != current_resname:
if current_het:
het_atoms.append(current_het)
current_het = [(x, y, z)]
current_resname = resname
else:
current_het.append((x, y, z))
elif line.startswith("ATOM") or line.startswith("TER"):
if current_het:
het_atoms.append(current_het)
current_het = []
current_resname = ""
if current_het:
het_atoms.append(current_het)
if not het_atoms:
return None
largest = max(het_atoms, key=len)
cx = sum(a[0] for a in largest) / len(largest)
cy = sum(a[1] for a in largest) / len(largest)
cz = sum(a[2] for a in largest) / len(largest)
return cx, cy, cz
def _find_protein_center(pdb_content: str) -> tuple[float, float, float]:
xs, ys, zs = [], [], []
for line in pdb_content.splitlines():
if line.startswith("ATOM") and len(line) >= 54:
try:
xs.append(float(line[30:38].strip()))
ys.append(float(line[38:46].strip()))
zs.append(float(line[46:54].strip()))
except ValueError:
continue
if not xs:
return 0.0, 0.0, 0.0
return sum(xs) / len(xs), sum(ys) / len(ys), sum(zs) / len(zs)
def _clean_protein(pdb_content: str) -> str:
"""Keep only ATOM records (protein), strip HETATM, waters, ANISOU, CONECT."""
lines: list[str] = []
for line in pdb_content.splitlines():
if line.startswith("ATOM") and len(line) >= 54:
lines.append(line)
elif line.startswith("TER"):
lines.append(line)
elif line.startswith("END"):
lines.append(line)
return "\n".join(lines)
def _parse_vina_pdbqt(pdbqt: str) -> list[dict[str, Any]]:
"""Parse Vina output PDBQT into individual pose dicts."""
models = re.split(r"^MODEL\s+(\d+)", pdbqt, flags=re.MULTILINE)
poses: list[dict[str, Any]] = []
current_atoms: list[dict[str, Any]] = []
current_model = 0
for chunk in models:
chunk = chunk.strip()
if chunk.isdigit():
current_model = int(chunk)
current_atoms = []
elif chunk and current_model > 0:
for line in chunk.splitlines():
if line.startswith("ATOM") or line.startswith("HETATM"):
try:
x = float(line[30:38].strip())
y = float(line[38:46].strip())
z = float(line[46:54].strip())
elem = line[76:78].strip()
current_atoms.append({"x": x, "y": y, "z": z, "element": elem})
except ValueError:
continue
if current_atoms:
energy_match = re.search(r"REMARK VINA RESULT:\s*([-\d.]+)", chunk)
poses.append({
"model": current_model,
"atoms": len(current_atoms),
"affinity": float(energy_match.group(1)) if energy_match else None,
})
current_atoms = []
return poses
class DockingTool(BaseTool):
name = "docking"
async def run(self, input: dict) -> dict:
pdb_id = input.get("pdb_id", "").strip().upper()
smiles = input.get("smiles", "").strip()
if not pdb_id or not smiles:
return {"error": "pdb_id and smiles are required"}
tmpdir = tempfile.mkdtemp(prefix="docking_")
try:
# 1. Fetch PDB
pdb_url = PDB_DOWNLOAD.format(pdb_id=pdb_id)
async with httpx.AsyncClient(timeout=30) as client:
r = await client.get(pdb_url)
if r.status_code != 200:
return {"error": f"PDB {pdb_id} not found at RCSB"}
pdb_content = r.text
pdb_path = os.path.join(tmpdir, "protein.pdb")
with open(pdb_path, "w") as f:
f.write(pdb_content)
# 2. Clean protein (strip waters, heteroatoms)
cleaned = _clean_protein(pdb_content)
clean_path = os.path.join(tmpdir, "cleaned.pdb")
with open(clean_path, "w") as f:
f.write(cleaned)
# 3. Convert protein to PDBQT via obabel
protein_pdbqt = os.path.join(tmpdir, "protein.pdbqt")
proc = await asyncio.create_subprocess_exec(
"obabel", clean_path, "-O", protein_pdbqt, "-xr",
stdout=asyncio.subprocess.PIPE,
stderr=asyncio.subprocess.PIPE,
)
_, stderr = await proc.communicate()
if proc.returncode != 0 or not os.path.exists(protein_pdbqt):
err = stderr.decode() if stderr else "obabel failed"
return {"error": f"Protein PDBQT preparation failed: {err}"}
# 4. Convert SMILES to 3D PDBQT via obabel
ligand_pdbqt = os.path.join(tmpdir, "ligand.pdbqt")
proc = await asyncio.create_subprocess_exec(
"obabel", f"-:{smiles}", "-O", ligand_pdbqt, "--gen3d", "-h",
stdout=asyncio.subprocess.PIPE,
stderr=asyncio.subprocess.PIPE,
)
_, stderr = await proc.communicate()
if proc.returncode != 0 or not os.path.exists(ligand_pdbqt):
err = stderr.decode() if stderr else "obabel failed"
return {"error": f"Ligand PDBQT preparation failed: {err}"}
# 5. Determine binding site box
center = _find_ligand_center(pdb_content)
if center:
cx, cy, cz = center
sx = sy = sz = 20
else:
cx, cy, cz = _find_protein_center(pdb_content)
sx = sy = sz = 30
# 6. Run Vina
out_pdbqt = os.path.join(tmpdir, "out.pdbqt")
vina_cmd = await asyncio.create_subprocess_exec(
VINA_CMD,
"--receptor", protein_pdbqt,
"--ligand", ligand_pdbqt,
"--out", out_pdbqt,
"--center_x", str(cx),
"--center_y", str(cy),
"--center_z", str(cz),
"--size_x", str(sx),
"--size_y", str(sy),
"--size_z", str(sz),
"--exhaustiveness", "8",
"--num_modes", "9",
stdout=asyncio.subprocess.PIPE,
stderr=asyncio.subprocess.PIPE,
)
try:
stdout, stderr = await asyncio.wait_for(vina_cmd.communicate(), timeout=600)
except asyncio.TimeoutError:
vina_cmd.kill()
await vina_cmd.communicate()
return {"error": "Docking timed out after 10 minutes"}
if vina_cmd.returncode != 0 or not os.path.exists(out_pdbqt):
err = stderr.decode("utf-8", errors="replace")[:500] if stderr else ""
return {"error": f"Vina failed (exit {vina_cmd.returncode}): {err}"}
# 7. Parse results
with open(out_pdbqt) as f:
out_content = f.read()
poses = _parse_vina_pdbqt(out_content)
log = stdout.decode() if stdout else ""
return {
"pdb_id": pdb_id,
"smiles": smiles,
"poses": poses,
"num_poses": len(poses),
"box_center": {"x": cx, "y": cy, "z": cz},
"box_size": {"x": sx, "y": sy, "z": sz},
"vina_log": log[:2000],
"from_cache": False,
}
except Exception as e:
logger.exception("Docking run failed")
return {"error": f"Docking failed: {e}"}
finally:
shutil.rmtree(tmpdir, ignore_errors=True)
|