Structure Identification

Classify local atomic environments and provide the structural labels many other plugins build on.

Overview

Structure Identification is one of the foundational plugins in the whole ecosystem. On its own, it classifies atoms by local structure. But it also acts as a supporting layer for other workflows such as OpenDXA and Grain Segmentation, which rely on those classifications and, in PTM mode, on orientation-related data as well.

That is why this plugin is worth understanding even if you rarely run it in isolation.

Parameters

Parameter	Type	Default	Description
`mode`	`string`	`CNA`	Identification algorithm. Options exposed in the product are `CNA`, `PTM`, and `DIAMOND`.
`rmsd`	`float`	`0.1`	RMSD threshold for PTM classification. Only applies in PTM mode.

Practical notes

PTM is the richest mode when you need orientations and correspondences.
CNA is the lighter structural screening path.
DIAMOND appears as an exposed mode, but the current implementation is not as cleanly separated internally as the UI may suggest.
The plugin also emits grouped atoms in a separate _atoms.msgpack file.

Output

`{outputBase}_structure_identification.msgpack`

type Vec3 = [float, float, float];
type Quaternion = [float, float, float, float];

interface PerAtomProperties {
    id: int;
    structure_type: int;
    structure_name: string;
    pos: Vec3;
}

interface PTMPerAtomProperties {
    id: int;
    structure_type: int;
    pos: Vec3;
    correspondence: int;
    orientation: Quaternion;
}

interface StructureIdentificationOutput {
    main_listing: {
        total_atoms: int;
        analysis_method: string;
        [key: `${string}_count`]: int;
        [key: `${string}_percentage`]: float;
        total_identified: int;
        total_unidentified: int;
        identification_rate: float;
        unique_structure_types: int;
    };
    "per-atom-properties": PerAtomProperties[] | PTMPerAtomProperties[];
}

`{outputBase}_atoms.msgpack`

Atoms grouped by structure name for visualization and downstream use.