Skip to main content

embed_diverse

Function embed_diverse 

Source
pub fn embed_diverse(
    smiles: &str,
    n_conformers: usize,
    rmsd_cutoff: f64,
    base_seed: u64,
) -> Vec<ConformerResult>
Expand description

Generate multiple conformers for a SMILES and filter by Butina RMSD clustering.

Generates n_conformers embeddings with different seeds, then clusters the successful ones by RMSD and returns only the cluster centroids (diverse set).

§Arguments

  • smiles: SMILES string
  • n_conformers: number of embedding attempts (different seeds)
  • rmsd_cutoff: RMSD threshold for clustering (Å), typically 1.0
  • base_seed: base seed for reproducibility (seeds = base_seed..base_seed+n_conformers)