Analyzing & Modifying Structures¶

Tip

Simmate's toolkit is built on top of PyMatGen (for crystals) and RDKit (for molecules). This tutorial serves as a guide to using these packages through a simplified, unified interface.

Quickstart¶

1. Crystals (Structures)¶

Load a structure from a file (like the POSCAR from the previous tutorial) and access its properties:

from simmate.toolkit import Structure

# Load from a file
structure = Structure.from_file("POSCAR")

# Access properties
print(f"Density: {structure.density}")
print(f"Volume: {structure.lattice.volume}")
print(f"Formula: {structure.composition.reduced_formula}")

# Modify and export
structure.make_supercell([2,2,2])
structure.to(filename="NaCl_supercell.cif", fmt="cif")

2. Molecules¶

Load a molecule from a SMILES string and explore its features:

from simmate.toolkit import Molecule

# Load from a SMILES string (e.g., Caffeine)
molecule = Molecule.from_smiles("CN1C=NC2=C1C(=O)N(C(=O)N2C)C")

# Access properties
print(f"Molecular Weight: {molecule.molecular_weight}")
print(f"Number of Rings: {molecule.num_rings}")
print(f"LogP: {molecule.log_p_rdkit}")

# Export to a file
molecule.to_sdf_file("caffeine.sdf")

Extra Examples¶

Looking for advanced features? Simmate incorporates many analytical tools directly into our toolkit, but even more are available through the underlying packages.

Random Structure Creation¶

Creating a random structure from a spacegroup and composition:

from simmate.toolkit import Composition
from simmate.toolkit.creators import RandomSymStructure

composition = Composition("Ca2N")
creator = RandomSymStructure(composition)

structure = creator.create_structure(spacegroup=166)

Fingerprints¶

Fingerprints are useful for comparing structures or molecules and creating machine-learning inputs.

CrystalsMolecules

# Using MatMiner (preinstalled with simmate)
from matminer.featurizers.structure.rdf import RadialDistributionFunction

rdf_analyzer = RadialDistributionFunction(bin_size=0.1)
rdf = rdf_analyzer.featurize(structure)

# Built-in to Simmate's Molecule class
fingerprint = molecule.fingerprint

Structure/Molecule Matching¶

Check if two objects are equivalent:

CrystalsMolecules

from pymatgen.analysis.structure_matcher import StructureMatcher

matcher = StructureMatcher()
is_matching = matcher.fit(structure1, structure2)

# Molecules can be compared directly
is_matching = (molecule1 == molecule2)