Expand description
Input builder types for ESMFold2 structure prediction.
This module provides typed builder structs for constructing the
StructurePredictionInput required by the ESMFold2 API. It mirrors
the Python esm.models.esmfold2 SDK, supporting protein chains,
DNA chains with optional chemical modifications, and small-molecule
ligands specified by CCD (Chemical Component Dictionary) codes.
§Example
use ferritin_plms::esmfold2::input_types::{
DNAInput, LigandInput, ProteinInput, StructurePredictionInput,
};
let protein = ProteinInput::new("A", "MKTAYIAK").unwrap();
let ligand = LigandInput::from_ccd("L", "ATP");
let input = StructurePredictionInput::new()
.add_protein(protein)
.add_ligand(ligand);
assert_eq!(input.num_chains(), 2);Structs§
- DNAInput
- A single DNA chain, optionally carrying chemical modifications.
- Ligand
Input - A small-molecule ligand described by one or more CCD codes.
- Modification
- A chemical modification at a specific position within a nucleic-acid chain.
- Protein
Input - A single protein chain for structure prediction.
- Structure
Prediction Input - The top-level input for an ESMFold2 structure prediction request.
Enums§
- Chain
Input - One chain in a multi-entity structure prediction request.
Functions§
- validate_
dna_ sequence - Validate that every character in a DNA sequence belongs to the accepted alphabet.
- validate_
protein_ sequence - Validate that every character in a protein sequence belongs to the accepted alphabet.