Quickstart

Load, score, minimize, and write a protein structure on CPU or CUDA. The example expects a local file named 1ubq.pdb.

Install TMol

python -m pip install tmol --only-binary=tmol

This installs the CPU build. For CUDA, follow Installation and change the device below to "cuda".

Load and score

import torch

from tmol.io import pose_stack_from_pdb
from tmol.score import beta2016_score_function

device = torch.device("cpu")
pose_stack = pose_stack_from_pdb("1ubq.pdb", device)

sfxn = beta2016_score_function(device)
scorer = sfxn.render_whole_pose_scoring_module(pose_stack)
score = scorer(pose_stack.coords)

print(score)

score contains one weighted total per pose.

Minimize and write

from tmol.optimization import run_cart_min

minimized = run_cart_min(pose_stack, sfxn)
rescored = scorer(minimized.coords)
print(rescored)

To write the result:

from tmol.io import write_pose_stack_pdb

write_pose_stack_pdb(minimized, "minimized.pdb")

Protein-ligand input

Load mmCIF with bond information for ligand preparation:

import biotite.structure as struc
import biotite.structure.io

from tmol.database import ParameterDatabase
from tmol.io import atom_array_from_file, pose_stack_from_biotite

structure = atom_array_from_file("complex.cif")

pose_stack, context = pose_stack_from_biotite(
    structure,
    device,
    prepare_ligands=True,
    param_db=ParameterDatabase.get_default(),
    return_context=True,
)

sfxn = beta2016_score_function(device, param_db=context.parameter_database)

Use the ligand-extended context.parameter_database when scoring a pose that contains freshly prepared ligands.

Continue with the structure tutorial or the scoring guide.