Expand description
chematic-fp — molecular fingerprints for chematic.
Provides:
- ECFP (Extended Connectivity Fingerprints): Morgan-algorithm-based circular fingerprints using FNV-1a hashing for reproducibility.
- MACCS 166-bit structural keys: SMARTS-based structural key fingerprints.
- Topological path fingerprints: DFS path enumeration up to a configurable length.
- AtomPair fingerprints: atom-pair encoding with topological distances.
- Topological Torsion fingerprints: four-atom path encoding.
- C-Series Reaction fingerprints: Chemical transformation encoding (Phase 1)
- MHFP/SECFP: MinHash fingerprints for fast approximate similarity searching (Phase 1)
- ERG: Extended Reduced Graph fingerprints for functional group-based similarity (Phase 1)
Re-exports§
pub use atom_pair::atom_pair_fp;pub use atom_pair::torsion_fp;pub use avalon::AvalonConfig;pub use avalon::avalon_fp;pub use avalon::avalon_fp_with_config;pub use avalon::tanimoto_avalon;pub use bitvec::BitVec2048;pub use bitvec::BitVecN;pub use bulk::tanimoto_matrix;pub use bulk::tanimoto_slice;pub use bulk::top_k_similar;pub use ecfp::EcfpConfig;pub use ecfp::ecfp;pub use ecfp::ecfp_with_bitinfo;pub use ecfp::ecfp4;pub use ecfp::ecfp6;pub use ecfp::morgan_fp_counts;pub use ecfp::tanimoto_ecfp4;pub use erg::ERG_VEC_LEN;pub use erg::ErgAtomType;pub use erg::ErgBondType;pub use erg::ErgConfig;pub use erg::ErgFingerprint;pub use erg::cosine_erg_vec;pub use erg::erg;pub use erg::erg_extended;pub use erg::erg_vec;pub use erg::erg_with_config;pub use erg::tanimoto_erg;pub use erg::tanimoto_erg_vec;pub use fcfp::fcfp;pub use fcfp::fcfp_with_bitinfo;pub use fcfp::fcfp4;pub use fcfp::fcfp6;pub use fcfp::tanimoto_fcfp4;pub use hdf::HdfConfig;pub use hdf::HdfFp;pub use hdf::cosine_hdf;pub use hdf::hdf;pub use hdf::hdf_default;pub use layered::layered_fp;pub use layered::layered_fp_by_layer;pub use layered::tanimoto_layered;pub use lsh::MhfpLshIndex;pub use maccs::maccs;pub use map4::Map4Config;pub use map4::map4;pub use map4::map4_default;pub use map4::tanimoto_map4;pub use mhfp::MhfpConfig;pub use mhfp::MhfpFingerprint;pub use mhfp::mhfp;pub use mhfp::mhfp_128;pub use mhfp::mhfp_with_config;pub use mhfp::tanimoto_mhfp;pub use path::RdkitPathConfig;pub use path::rdkit_path_fp;pub use path::rdkit_path_fp_with_config;pub use path::tanimoto_rdkit_path;pub use pattern::pattern_fp;pub use pattern::tanimoto_pattern;pub use pharmacophore_fp::pharmacophore_feature_counts;pub use pharmacophore_fp::pharmacophore_fp_2d;pub use pharmacophore_fp::tanimoto_pharmacophore_2d;pub use reaction_fp::ReactionFingerprint;pub use reaction_fp::ReactionFpConfig;pub use reaction_fp::reaction_fp;pub use reaction_fp::reaction_fp_ecfp4;pub use reaction_fp::reaction_fp_with_config;pub use reaction_fp::tanimoto_reaction_fp;pub use search::FpType;pub use search::nearest_neighbors;pub use search::nearest_neighbors_from_fp;pub use topo_path::TopoPathConfig;pub use topo_path::tanimoto_topo_path;pub use topo_path::topo_path;
Modules§
- atom_
pair - Atom-Pair and Topological Torsion fingerprints.
- avalon
- Avalon-style structural fingerprint.
- bitvec
- Fixed-size 2048-bit bitvector backed by 32 × u64 words.
- bulk
- Bulk Tanimoto computation over BitVec2048 fingerprint slices.
- ecfp
- ECFP (Extended Connectivity Fingerprints) based on the Morgan algorithm.
- erg
- ERG (Extended Reduced Graph) fingerprints — bespoke 2048-bit and 315-dim float variants.
- fcfp
- FCFP (Feature-Class Fingerprints) — pharmacophore-based circular fingerprints.
- hdf
- Hyper-Dimensional Fingerprints (HDF) for molecules.
- layered
- Layered fingerprint: RDKit-like structural layers encoding atom features.
- lsh
- MinHash LSH (Locality Sensitive Hashing) index for fast approximate Jaccard similarity search over MHFP fingerprints.
- maccs
- MACCS 166-bit structural keys fingerprint.
- map4
- MAP4 — MinHashed Atom-Pair Fingerprint (Minervini et al. 2020)
- mhfp
- MHFP (MinHash Fingerprint) — Circular SMILES MinHash
- path
- v0.1.92: RDKit-compatible path fingerprints with bond type inclusion.
- pattern
- Pattern fingerprints — substructure feature hashing.
- pharmacophore_
fp - Pharmacophore fingerprints for virtual screening and similarity search.
- reaction_
fp - C-Series Phase 1: Reaction fingerprints for chemical transformation encoding.
- search
- Similarity-based nearest-neighbour search over a molecule database.
- topo_
path - Topological path fingerprints.