Core Workflow

This tutorial covers the core workflow in automating NMR predictions via CASTEP, with a section on how it can be extended to automated ML training in a wider autoplex workflow (such as the RSS workflow).

Overview

Given a list of pymatgen structures (usually given in a .xyz file) and user-defined parameters, CASTEP is called with task: magres (see the CASTEP setup for how autoplex runs CASTEP) and returns .magres and .castep files, as well as the magnetic shielding (MS) and electric field gradient (EFG) tensors for each atom in each structure.

Example workflow

First we define our .param and .cell parameters using CastepMagresSetGenerator:

from autoplex.misc.castep.utils import CastepMagresSetGenerator

input_set_generator = CastepMagresSetGenerator(
    use_efg=True,
    user_param_settings={"xc_functional": "PBE", "cut_off_energy": 900.0},
    user_cell_settings={
        "species_pot": [
            ("O", "2|1.1|17|20|23|20:21(qc=8)"),
            ("Si", "3|1.8|5|6|7|30:31:32"),
        ],
        "kpoint_mp_spacing": 0.05,
    },
)

All allowed CASTEP keywords and definitions for setting up .param and .cell files can be found in the autoplex package at: /yourpath/autoplex/misc/castep/castep_keywords.json.

use_efg determines whether EFG tensors are calculated as well, if set to False it only calculates MS tensors.

Next, using input_set_generator we create a CastepMagresMaker:

from autoplex.misc.castep.jobs import CastepMagresMaker

magres_maker = CastepMagresMaker(
    input_set_generator=input_set_generator
)

Then we load our structures from e.g. structures.xyz:

from pymatgen.io.ase import AseAtomsAdaptor
from ase.io import read

structures = [AseAtomsAdaptor.get_structure(structure) for structure in read("structures.xyz", ":")]

This creates an array of pymatgen Structure objects which can then be passed into CastepMagresFlowMaker, which creates the Flow to be run locally or submitted to HPC (see the jobflow-remote setup):

from jobflow import run_locally

from autoplex.misc.castep.flows import CastepMagresFlowMaker

flow = CastepMagresFlowMaker(magres_maker=magres_maker).make(structures)
run_locally(flow, create_folders=True)  # or jobflow-remote's submit_flow

Each job returns a TaskDoc. The tensors are in output.ms_tensor (ppm) and output.efg_tensor (atomic units), with one 3×3 tensor per atom, ordered like output.structure. This may differ from the input order, because CASTEP groups atoms by element. The compressed castep.castep.gz and castep.magres.gz files stay in the job’s CASTEP/ folder.

Next steps

The core workflow should be ideally extended to automatically train ML models for NMR prediction (Ben Mahmoud et al, J. Chem. Phys. 163, 024118 (2025)) in a similar workflow to RssMaker (see the RSS workflow). The outputted .magres files could be used in the same way for labelling the dataset, however there will be differences to RssMaker in data processing, generation and sampling as similiar energy structures can have wildly different NMR parameters.